You are not authenticated, login.
text: sort by
tags: modified
type: chronology
{847} is owned by tlh24.{721} is owned by tlh24 misha.
[0] Gage GJ, Ludwig KA, Otto KJ, Ionides EL, Kipke DR, Naive coadaptive cortical control.J Neural Eng 2:2, 52-63 (2005 Jun)

[0] Jackson A, Mavoori J, Fetz EE, Correlations between the same motor cortex cells and arm muscles during a trained task, free behavior, and natural sleep in the macaque monkey.J Neurophysiol 97:1, 360-74 (2007 Jan)

[0] Schmidt EM, McIntosh JS, Durelli L, Bak MJ, Fine control of operantly conditioned firing patterns of cortical neurons.Exp Neurol 61:2, 349-69 (1978 Sep 1)[1] Serruya MD, Hatsopoulos NG, Paninski L, Fellows MR, Donoghue JP, Instant neural control of a movement signal.Nature 416:6877, 141-2 (2002 Mar 14)[2] Fetz EE, Operant conditioning of cortical unit activity.Science 163:870, 955-8 (1969 Feb 28)[3] Fetz EE, Finocchio DV, Operant conditioning of specific patterns of neural and muscular activity.Science 174:7, 431-5 (1971 Oct 22)[4] Fetz EE, Finocchio DV, Operant conditioning of isolated activity in specific muscles and precentral cells.Brain Res 40:1, 19-23 (1972 May 12)[5] Fetz EE, Baker MA, Operantly conditioned patterns on precentral unit activity and correlated responses in adjacent cells and contralateral muscles.J Neurophysiol 36:2, 179-204 (1973 Mar)

[0] Hochberg LR, Serruya MD, Friehs GM, Mukand JA, Saleh M, Caplan AH, Branner A, Chen D, Penn RD, Donoghue JP, Neuronal ensemble control of prosthetic devices by a human with tetraplegia.Nature 442:7099, 164-71 (2006 Jul 13)

[0] Mehring C, Rickert J, Vaadia E, Cardosa de Oliveira S, Aertsen A, Rotter S, Inference of hand movements from local field potentials in monkey motor cortex.Nat Neurosci 6:12, 1253-4 (2003 Dec)

[0] Rousche PJ, Normann RA, Chronic recording capability of the Utah Intracortical Electrode Array in cat sensory cortex.J Neurosci Methods 82:1, 1-15 (1998 Jul 1)

[0] Santhanam G, Ryu SI, Yu BM, Afshar A, Shenoy KV, A high-performance brain-computer interface.Nature 442:7099, 195-8 (2006 Jul 13)[1] Shenoy KV, Meeker D, Cao S, Kureshi SA, Pesaran B, Buneo CA, Batista AP, Mitra PP, Burdick JW, Andersen RA, Neural prosthetic control signals from plan activity.Neuroreport 14:4, 591-6 (2003 Mar 24)

[0] Musallam S, Corneil BD, Greger B, Scherberger H, Andersen RA, Cognitive control signals for neural prosthetics.Science 305:5681, 258-62 (2004 Jul 9)

[0] Carmena JM, Lebedev MA, Crist RE, O'Doherty JE, Santucci DM, Dimitrov DF, Patil PG, Henriquez CS, Nicolelis MA, Learning to control a brain-machine interface for reaching and grasping by primates.PLoS Biol 1:2, E42 (2003 Nov)

[0] Taylor DM, Tillery SI, Schwartz AB, Direct cortical control of 3D neuroprosthetic devices.Science 296:5574, 1829-32 (2002 Jun 7)

[0] Serruya M, Hatsopoulos N, Fellows M, Paninski L, Donoghue J, Robustness of neuroprosthetic decoding algorithms.Biol Cybern 88:3, 219-28 (2003 Mar)

[0] Wessberg J, Stambaugh CR, Kralik JD, Beck PD, Laubach M, Chapin JK, Kim J, Biggs SJ, Srinivasan MA, Nicolelis MA, Real-time prediction of hand trajectory by ensembles of cortical neurons in primates.Nature 408:6810, 361-5 (2000 Nov 16)

[0] Kennedy PR, Bakay RA, Moore MM, Adams K, Goldwaithe J, Direct control of a computer from the human central nervous system.IEEE Trans Rehabil Eng 8:2, 198-202 (2000 Jun)[1] Kennedy PR, Mirra SS, Bakay RA, The cone electrode: ultrastructural studies following long-term recording in rat and monkey cortex.Neurosci Lett 142:1, 89-94 (1992 Aug 3)

[0] Fetz EE, Baker MA, Operantly conditioned patterns on precentral unit activity and correlated responses in adjacent cells and contralateral muscles.J Neurophysiol 36:2, 179-204 (1973 Mar)

[0] Fetz EE, Operant conditioning of cortical unit activity.Science 163:870, 955-8 (1969 Feb 28)[1] Fetz EE, Finocchio DV, Operant conditioning of specific patterns of neural and muscular activity.Science 174:7, 431-5 (1971 Oct 22)[2] Fetz EE, Finocchio DV, Operant conditioning of isolated activity in specific muscles and precentral cells.Brain Res 40:1, 19-23 (1972 May 12)

[0] Ashe J, Force and the motor cortex.Behav Brain Res 87:2, 255-69 (1997 Sep)[1] Cabel DW, Cisek P, Scott SH, Neural activity in primary motor cortex related to mechanical loads applied to the shoulder and elbow during a postural task.J Neurophysiol 86:4, 2102-8 (2001 Oct)[2] Cheney PD, Fetz EE, Functional classes of primate corticomotoneuronal cells and their relation to active force.J Neurophysiol 44:4, 773-91 (1980 Oct)[3] Evarts EV, Relation of pyramidal tract activity to force exerted during voluntary movement.J Neurophysiol 31:1, 14-27 (1968 Jan)[4] Evarts EV, Activity of pyramidal tract neurons during postural fixation.J Neurophysiol 32:3, 375-85 (1969 May)[5] Humphrey DR, Schmidt EM, Thompson WD, Predicting measures of motor performance from multiple cortical spike trains.Science 170:959, 758-62 (1970 Nov 13)[6] Thach WT, Correlation of neural discharge with pattern and force of muscular activity, joint position, and direction of intended next movement in motor cortex and cerebellum.J Neurophysiol 41:3, 654-76 (1978 May)[7] Wetts R, Kalaska JF, Smith AM, Cerebellar nuclear cell activity during antagonist cocontraction and reciprocal inhibition of forearm muscles.J Neurophysiol 54:2, 231-44 (1985 Aug)[8] Georgopoulos AP, Ashe J, Smyrnis N, Taira M, The motor cortex and the coding of force.Science 256:5064, 1692-5 (1992 Jun 19)[9] Kalaska JF, Cohen DA, Hyde ML, Prud'homme M, A comparison of movement direction-related versus load direction-related activity in primate motor cortex, using a two-dimensional reaching task.J Neurosci 9:6, 2080-102 (1989 Jun)[10] Li CS, Padoa-Schioppa C, Bizzi E, Neuronal correlates of motor performance and motor learning in the primary motor cortex of monkeys adapting to an external force field.Neuron 30:2, 593-607 (2001 May)[11] Sergio LE, Kalaska JF, Systematic changes in directional tuning of motor cortex cell activity with hand location in the workspace during generation of static isometric forces in constant spatial directions.J Neurophysiol 78:2, 1170-4 (1997 Aug)[12] Sergio LE, Kalaska JF, Systematic changes in motor cortex cell activity with arm posture during directional isometric force generation.J Neurophysiol 89:1, 212-28 (2003 Jan)[13] Taira M, Boline J, Smyrnis N, Georgopoulos AP, Ashe J, On the relations between single cell activity in the motor cortex and the direction and magnitude of three-dimensional static isometric force.Exp Brain Res 109:3, 367-76 (1996 Jun)[14] Maier MA, Bennett KM, Hepp-Reymond MC, Lemon RN, Contribution of the monkey corticomotoneuronal system to the control of force in precision grip.J Neurophysiol 69:3, 772-85 (1993 Mar)[15] Hepp-Reymond M, Kirkpatrick-Tanner M, Gabernet L, Qi HX, Weber B, Context-dependent force coding in motor and premotor cortical areas.Exp Brain Res 128:1-2, 123-33 (1999 Sep)[16] Smith AM, Hepp-Reymond MC, Wyss UR, Relation of activity in precentral cortical neurons to force and rate of force change during isometric contractions of finger muscles.Exp Brain Res 23:3, 315-32 (1975 Sep 29)[17] Cooke JD, Brown SH, Movement-related phasic muscle activation. II. Generation and functional role of the triphasic pattern.J Neurophysiol 63:3, 465-72 (1990 Mar)[18] Almeida GL, Hong DA, Corcos D, Gottlieb GL, Organizing principles for voluntary movement: extending single-joint rules.J Neurophysiol 74:4, 1374-81 (1995 Oct)[19] Gottlieb GL, Latash ML, Corcos DM, Liubinskas TJ, Agarwal GC, Organizing principles for single joint movements: V. Agonist-antagonist interactions.J Neurophysiol 67:6, 1417-27 (1992 Jun)[20] Corcos DM, Agarwal GC, Flaherty BP, Gottlieb GL, Organizing principles for single-joint movements. IV. Implications for isometric contractions.J Neurophysiol 64:3, 1033-42 (1990 Sep)[21] Gottlieb GL, Corcos DM, Agarwal GC, Latash ML, Organizing principles for single joint movements. III. Speed-insensitive strategy as a default.J Neurophysiol 63:3, 625-36 (1990 Mar)[22] Corcos DM, Gottlieb GL, Agarwal GC, Organizing principles for single-joint movements. II. A speed-sensitive strategy.J Neurophysiol 62:2, 358-68 (1989 Aug)[23] Gottlieb GL, Corcos DM, Agarwal GC, Organizing principles for single-joint movements. I. A speed-insensitive strategy.J Neurophysiol 62:2, 342-57 (1989 Aug)[24] Ghez C, Gordon J, Trajectory control in targeted force impulses. I. Role of opposing muscles.Exp Brain Res 67:2, 225-40 (1987)[25] Sainburg RL, Ghez C, Kalakanis D, Intersegmental dynamics are controlled by sequential anticipatory, error correction, and postural mechanisms.J Neurophysiol 81:3, 1045-56 (1999 Mar)[26] Georgopoulos AP, Kalaska JF, Caminiti R, Massey JT, On the relations between the direction of two-dimensional arm movements and cell discharge in primate motor cortex.J Neurosci 2:11, 1527-37 (1982 Nov)[27] Ashe J, Georgopoulos AP, Movement parameters and neural activity in motor cortex and area 5.Cereb Cortex 4:6, 590-600 (1994 Nov-Dec)[28] Caminiti R, Johnson PB, Urbano A, Making arm movements within different parts of space: dynamic aspects in the primate motor cortex.J Neurosci 10:7, 2039-58 (1990 Jul)[29] Caminiti R, Johnson PB, Galli C, Ferraina S, Burnod Y, Making arm movements within different parts of space: the premotor and motor cortical representation of a coordinate system for reaching to visual targets.J Neurosci 11:5, 1182-97 (1991 May)[30] Matsuzaka Y, Picard N, Strick PL, Skill representation in the primary motor cortex after long-term practice.J Neurophysiol 97:2, 1819-32 (2007 Feb)[31] Fu QG, Suarez JI, Ebner TJ, Neuronal specification of direction and distance during reaching movements in the superior precentral premotor area and primary motor cortex of monkeys.J Neurophysiol 70:5, 2097-116 (1993 Nov)

[0] BASMAJIAN JV, Control and training of individual motor units.Science 141no Issue 440-1 (1963 Aug 2)

[0] Schmidt EM, Single neuron recording from motor cortex as a possible source of signals for control of external devices.Ann Biomed Eng 8:4-6, 339-49 (1980)[1] Schmidt EM, McIntosh JS, Durelli L, Bak MJ, Fine control of operantly conditioned firing patterns of cortical neurons.Exp Neurol 61:2, 349-69 (1978 Sep 1)[2] Salcman M, Bak MJ, A new chronic recording intracortical microelectrode.Med Biol Eng 14:1, 42-50 (1976 Jan)

[0] Ferrari PF, Rozzi S, Fogassi L, Mirror neurons responding to observation of actions made with tools in monkey ventral premotor cortex.J Cogn Neurosci 17:2, 212-26 (2005 Feb)[1] Maravita A, Iriki A, Tools for the body (schema).Trends Cogn Sci 8:2, 79-86 (2004 Feb)[2] Sanchez J, Principe J, Carmena J, Lebedev M, Nicolelis MA, Simultaneus prediction of four kinematic variables for a brain-machine interface using a single recurrent neural network.Conf Proc IEEE Eng Med Biol Soc 7no Issue 5321-4 (2004)[3] Wood F, Fellows M, Donoghue J, Black M, Automatic spike sorting for neural decoding.Conf Proc IEEE Eng Med Biol Soc 6no Issue 4009-12 (2004)[4] Mehring C, Rickert J, Vaadia E, Cardosa de Oliveira S, Aertsen A, Rotter S, Inference of hand movements from local field potentials in monkey motor cortex.Nat Neurosci 6:12, 1253-4 (2003 Dec)[5] Won DS, Wolf PD, A simulation study of information transmission by multi-unit microelectrode recordings.Network 15:1, 29-44 (2004 Feb)[6] Schmidt EM, Single neuron recording from motor cortex as a possible source of signals for control of external devices.Ann Biomed Eng 8:4-6, 339-49 (1980)[7] Salcman M, Bak MJ, A new chronic recording intracortical microelectrode.Med Biol Eng 14:1, 42-50 (1976 Jan)[8] Patil PG, Carmena JM, Nicolelis MA, Turner DA, Ensemble recordings of human subcortical neurons as a source of motor control signals for a brain-machine interface.Neurosurgery 55:1, 27-35; discussion 35-8 (2004 Jul)[9] Santhanam G, Ryu SI, Yu BM, Afshar A, Shenoy KV, A high-performance brain-computer interface.Nature 442:7099, 195-8 (2006 Jul 13)[10] Brockwell AE, Rojas AL, Kass RE, Recursive bayesian decoding of motor cortical signals by particle filtering.J Neurophysiol 91:4, 1899-907 (2004 Apr)[11] Marzullo TC, Miller CR, Kipke DR, Suitability of the cingulate cortex for neural control.IEEE Trans Neural Syst Rehabil Eng 14:4, 401-9 (2006 Dec)[12] Jackson A, Mavoori J, Fetz EE, Long-term motor cortex plasticity induced by an electronic neural implant.Nature 444:7115, 56-60 (2006 Nov 2)

[0] Carmena JM, Lebedev MA, Henriquez CS, Nicolelis MA, Stable ensemble performance with single-neuron variability during reaching movements in primates.J Neurosci 25:46, 10712-6 (2005 Nov 16)

[1] Obeid I, Nicolelis MA, Wolf PD, A low power multichannel analog front end for portable neural signal recordings.J Neurosci Methods 133:1-2, 27-32 (2004 Feb 15)

[0] Kennedy PR, Mirra SS, Bakay RA, The cone electrode: ultrastructural studies following long-term recording in rat and monkey cortex.Neurosci Lett 142:1, 89-94 (1992 Aug 3)

[0] Jackson A, Mavoori J, Fetz EE, Long-term motor cortex plasticity induced by an electronic neural implant.Nature 444:7115, 56-60 (2006 Nov 2)

[0] Patil PG, Carmena JM, Nicolelis MA, Turner DA, Ensemble recordings of human subcortical neurons as a source of motor control signals for a brain-machine interface.Neurosurgery 55:1, 27-35; discussion 35-8 (2004 Jul)

[0] Won DS, Wolf PD, A simulation study of information transmission by multi-unit microelectrode recordings.Network 15:1, 29-44 (2004 Feb)

[0] Radhakrishnan SM, Baker SN, Jackson A, Learning a novel myoelectric-controlled interface task.J Neurophysiol no Volume no Issue no Pages (2008 Jul 30)

[0] Is this the bionic man?Nature 442:7099, 109 (2006 Jul 13)[1] Hochberg LR, Serruya MD, Friehs GM, Mukand JA, Saleh M, Caplan AH, Branner A, Chen D, Penn RD, Donoghue JP, Neuronal ensemble control of prosthetic devices by a human with tetraplegia.Nature 442:7099, 164-71 (2006 Jul 13)[2] Santhanam G, Ryu SI, Yu BM, Afshar A, Shenoy KV, A high-performance brain-computer interface.Nature 442:7099, 195-8 (2006 Jul 13)[3] Shenoy KV, Meeker D, Cao S, Kureshi SA, Pesaran B, Buneo CA, Batista AP, Mitra PP, Burdick JW, Andersen RA, Neural prosthetic control signals from plan activity.Neuroreport 14:4, 591-6 (2003 Mar 24)

[0] Narayanan NS, Kimchi EY, Laubach M, Redundancy and synergy of neuronal ensembles in motor cortex.J Neurosci 25:17, 4207-16 (2005 Apr 27)

[0] Serruya MD, Hatsopoulos NG, Paninski L, Fellows MR, Donoghue JP, Instant neural control of a movement signal.Nature 416:6877, 141-2 (2002 Mar 14)

[0] Fetz EE, Volitional control of neural activity: implications for brain-computer interfaces.J Physiol 579:Pt 3, 571-9 (2007 Mar 15)

[0] Ojakangas CL, Shaikhouni A, Friehs GM, Caplan AH, Serruya MD, Saleh M, Morris DS, Donoghue JP, Decoding movement intent from human premotor cortex neurons for neural prosthetic applications.J Clin Neurophysiol 23:6, 577-84 (2006 Dec)

[0] Humphrey DR, Schmidt EM, Thompson WD, Predicting measures of motor performance from multiple cortical spike trains.Science 170:959, 758-62 (1970 Nov 13)

[0] Birbaumer N, Cohen LG, Brain-computer interfaces: communication and restoration of movement in paralysis.J Physiol 579:Pt 3, 621-36 (2007 Mar 15)

[0] Wahnoun R, Helms Tillery S, He J, Neuron selection and visual training for population vector based cortical control.Conf Proc IEEE Eng Med Biol Soc 6no Issue 4607-10 (2004)[1] Wahnoun R, He J, Helms Tillery SI, Selection and parameterization of cortical neurons for neuroprosthetic control.J Neural Eng 3:2, 162-71 (2006 Jun)[2] Fetz EE, Operant conditioning of cortical unit activity.Science 163:870, 955-8 (1969 Feb 28)[3] Fetz EE, Finocchio DV, Operant conditioning of specific patterns of neural and muscular activity.Science 174:7, 431-5 (1971 Oct 22)[4] Fetz EE, Finocchio DV, Operant conditioning of isolated activity in specific muscles and precentral cells.Brain Res 40:1, 19-23 (1972 May 12)[5] Fetz EE, Baker MA, Operantly conditioned patterns on precentral unit activity and correlated responses in adjacent cells and contralateral muscles.J Neurophysiol 36:2, 179-204 (1973 Mar)[6] Humphrey DR, Schmidt EM, Thompson WD, Predicting measures of motor performance from multiple cortical spike trains.Science 170:959, 758-62 (1970 Nov 13)

[0] Wood F, Fellows M, Donoghue J, Black M, Automatic spike sorting for neural decoding.Conf Proc IEEE Eng Med Biol Soc 6no Issue 4009-12 (2004)

[0] Sanchez J, Principe J, Carmena J, Lebedev M, Nicolelis MA, Simultaneus prediction of four kinematic variables for a brain-machine interface using a single recurrent neural network.Conf Proc IEEE Eng Med Biol Soc 7no Issue 5321-4 (2004)

[0] Schwartz AB, Cortical neural prosthetics.Annu Rev Neurosci 27no Issue 487-507 (2004)[1] Carmena JM, Lebedev MA, Henriquez CS, Nicolelis MA, Stable ensemble performance with single-neuron variability during reaching movements in primates.J Neurosci 25:46, 10712-6 (2005 Nov 16)

[0] O'Neill DE, Corporate executive says era of the "bionic" man is now a scientific fact.Rev Fed Am Hosp 13:5-6, 22-3 (1980 Sep-Oct)[1] Marbach WD, Zabarsky M, Hoban P, Nelson C, Building the bionic man.Newsweek 100:2, 78-9 (1982 Jul 12)[2] Craelius W, The bionic man: restoring mobility.Science 295:5557, 1018-21 (2002 Feb 8)[3] Vogel G, Part man, part computer: researcher tests the limits.Science 295:5557, 1020 (2002 Feb 8)

[0] Kamitani Y, Tong F, Decoding the visual and subjective contents of the human brain.Nat Neurosci 8:5, 679-85 (2005 May)

[0] Brockwell AE, Rojas AL, Kass RE, Recursive bayesian decoding of motor cortical signals by particle filtering.J Neurophysiol 91:4, 1899-907 (2004 Apr)

[0] Marzullo TC, Miller CR, Kipke DR, Suitability of the cingulate cortex for neural control.IEEE Trans Neural Syst Rehabil Eng 14:4, 401-9 (2006 Dec)

hide / / print
ref: -2020 tags: Neuralink commentary BMI pigs date: 08-31-2020 18:01 gmt revision:1 [0] [head]

Neuralink progress update August 28 2020

Some commentary.

The good:

  • Ian hit the nail on the head @ 1:05:47. That is not a side-benefit -- that was the original and true purpose. Thank you.
  • The electronics, amplify / record / sort / stim ASIC, as well as interconnect all advance the state of the art in density, power efficiency, and capability. (I always liked higher sampling rates, but w/e)
  • Puck is an ideal form factor, again SOTA. 25mm diameter craniotomy should give plenty of space for 32 x 32-channel depth electrodes (say).
  • I would estimate that the high-density per electrode feed-through is also SOTA, but it might also be a non-hermetic pass-through via the thin-film (e.g. some water vapor diffusion along the length of the polyimide (if that polymer is being used)).
  • Robot looks nice dressed in those fancy robes. Also looks like there is a revolute joint along the coronal axis.
  • Stim on every channel is cool.
  • Pigs seem like an ethical substitute for monkeys.

The mixed:

  • Neurons are not wires.
  • $2000 outpatient neurosurgery?! Will need to address the ~3% complication rate for most neurosurgery.
  • Where is the monkey data? Does it not work in monkeys? Insufficient longevity or yield? Was it strategic to not mention any monkeys, to avoid bad PR or the wrath of PETA?
    • I can't imagine getting into humans without demonstrating both safety and effectiveness on monkeys. Pigs are fine for the safety part, but monkeys are the present standard for efficacy.
  • How long do the electrodes last in pigs? What is the recording quality? How stable are the traces?
    • Judging from the commentary, assume this is a electrode material problem? What does Neuralink do if they are not significantly different in yield and longevity than the Utah array? (The other problems might well be easier than this one.)
      • That said, a thousand channels of EMG should be sufficient for some of the intended applications (below).
    • It really remains to be seen how well the brain tolerates these somewhat-large somewhat-thin electrodes, what percentage of the brain is disrupted in the process of insertion, and how much of the disruption is transient / how much is irrecoverable.
    • Pig-snout somatosensory cortex is an unusual recording location, making comparison difficult, but what was shown seemed rather correlated (?) We'd have to read an actual scientific publication to evaluate.
  • This slide is deceptive, as not all the applications are equally .. applicable. You don't need an extracellular ephys device to solve these problems that "almost everyone" will encounter over the course of their lives.
    • Memory loss -- Probably better dealt with via cellular / biological therapies, or treating the causes (stroke, infection, inflammation, neuroendocrine or neuromodulatory disregulation)
    • Hearing loss -- Reasonable. Nice complement to improved cochlear implants too. (Maybe the Neuralink ASIC could be used for that, too).
      • With this and the other reasonable applications, best to keep in context that stereo EEG, which is fairly disruptive w/ large probes, is well tolerated in epilepsy patients. (It has unclear effect on IQ or memory, but still, the sewing machine should be less invasive.)
    • Blindness -- Reasonable. Mating the puck to a Second Sight style thin film would improve channel count dramatically, and be less invasive. Otherwise you have to sew into the calcarine fissure, destroying a fair bit of cortex in the process & possibly hitting an artery or sulcal vein.
    • Paralysis -- Absolutely. This application is well demonstrated, and the Neuralink device should be able to help SCI patients. Presumably this will occupy them for the next five years; other applications would be a distraction.
      • Being able to sew flexible electrodes into the spinal cord is a great application.
    • Depression -- Need deeper targets for this. Research to treat depression via basal ganglia stim is ongoing; no reason it could not be mated to the Neuralink puck + long electrodes.
    • Insomina -- I guess?
    • Extreme pain -- Simpler approaches are likely better, but sure?
    • Seizures -- Yes, but note that Neuropace burned through $250M and wasn't significantly better than sham surgery. Again, likely better dealt with biologically: recombinant ion channels, glial or interneuron stem cell therapy.
    • Anxiety -- maybe? Designer drugs seem safer. Or drugs + CBT. Elon likes root causes: spotlight on the structural ills of our society.
    • Addiction -- Yes. It seems possible to rewire the brain with the right record / stim strategy, via for example a combination of DBS and cortical recording. Social restructuring is again a better root-cause fix.
    • Strokes -- No, despite best efforts, the robot causes (small) strokes.
    • Brain Damage -- Insertion of electrodes causes brain damage. Again, better dealt with via cellular (e.g. stem cells) or biological approaches.
      • This, of course, will take time as our understanding of brain development is limited; the good thing is that sufficient guidance signals remain in the adult brain, so AFAIK it's possible. From his comments, seems Alan's attitude is more aligned with this.
    • Not really bad per-se, but right panel could be better. I assume this was a design decision trade-off between working distance, NA, illumination, and mechanical constraints.
    • Despite Elon's claims, there is always bleeding when you poke electrodes that large into the cortex; the capillary bed is too dense. Let's assume Elon meant 'macro' bleeding, which is true. At least the robot avoids visible vessels.
    • Predicting joint angles for cyclical behavior is not challenging; can be done with EMG or microphonic noise correlated to some part of the gait. Hence the request for monkey BMI data.
  • Given the risk, pretty much any of the "sci-fi" applications mentioned in response to dorky twitter comments can be better provided to neurologically normal people through electronics, without the risk of brain surgery.
  • Regarding sci-fi application linguistic telepathy:
    • First, agreed, clarifying thoughts into language takes effort. This is a mostly unavoidable and largely good task. Interfacing with the external world is a vital part of cognition; shortcutting it, in my estimation, will just lead to sloppy & half-formed ideas not worth communicating. The compression of thoughts into words (as lossy as it may be) is the primary way to make them discrete enough to be meaningful to both other people and yourself.
    • Secondly: speech (or again any of the many other forms of communication) is not that much slower than cognition. If it was, we'd have much larger vocabularies, much more complicated and meaning-conveying grammar, etc (Like Latin?). The limit is the average persons cognition and memory. I disagree with Elon's conceit.
  • Regarding visual telepathy, with sufficient recording capabilities, I see no reason why you couldn't have a video-out port on the brain. Difficult given the currently mostly unknown representation of higher-level visual cortices, but as Ian says, once you have a good oscilloscope, this can be deduced.
  • Regarding AI symbiosis @1:09:19; this logic is not entirely clear to me. AI is a tool that will automate & facilitate the production and translation of knowledge much the same way electricity etc automated & facilitated the production and transportation of physical goods. We will necessarily need to interface with it, but to the point that we are thoroughly modifying our own development & biology, those interfaces will likely be based on presently extant computer interfaces.
    • If we do start modifying the biological wiring structure of our brains, I can't imagine that there will many limits! (Outside hard metabolic limits that brain vasculature takes pains to allocate and optimize.)
    • So, I guess the central tenet might be vaguely ok if you allow that humans are presently symbiotic with cell phones. (A more realistic interpretation is that cell phones are tools, and maybe Google etc are the symbionts / parasites). This is arguably contributing to current political existential crises -- no need to look further. If you do look further, it's not clear that stabbing the brains of healthy individuals will help.
    • I find the MC to be slightly unctuous and ingratiating in a way appropriate for a video game company, but not for a medical device company. That, of course, is a judgement call & matter of taste. Yet, as this was partly a recruiting event ... you will find who you set the table for.

hide / / print
ref: Gage-2005.06 tags: naive coadaptive control Kalman filter Kipke audio BMI date: 09-13-2019 02:33 gmt revision:2 [1] [0] [head]

PMID-15928412[0] Naive coadaptive Control May 2005. see notes


hide / / print
ref: Jackson-2007.01 tags: Fetz neurochip sleep motor control BMI free behavior EMG date: 09-13-2019 02:21 gmt revision:4 [3] [2] [1] [0] [head]

PMID-17021028[0] Correlations Between the Same Motor Cortex Cells and Arm Muscles During a Trained Task, Free Behavior, and Natural Sleep in the Macaque Monkey

  • used their implanted "neurochip" recorder that recorded both EMG and neural activity. The neurochip buffers data and transmits via IR offline. It doesn't have all that much flash onboard - 16Mb.
    • used teflon-insulated 50um tungsten wires.
  • confirmed that there is a strong causal relationship, constant over the course of weeks, between motor cortex units and EMG activity.
    • some causal relationships between neural firing and EMG varied dependent on the task. Additive / multiplicative encoding?
  • this relationship was different at night, during REM sleep, though (?)
  • point out, as Todorov did, that Stereotyped motion imposes correlation between movement parameters, which could lead to spurrious relationships being mistaken for neural coding.
    • Experiments with naturalistic movement are essential for understanding innate, untrained neural control.
  • references {597} Suner et al 2005 as a previous study of long term cortical recordings. (utah probe)
  • during sleep, M1 cells exhibited a cyclical patter on quiescence followed by periods of elevated activity;
    • the cycle lasted 40-60 minutes;
    • EMG activity was seen at entrance and exit to the elevated activity period.
    • during periods of highest cortical activity, muscle activity was completely suppressed.
    • peak firing rates were above 100hz! (mean: 12-16hz).


hide / / print
ref: Schmidt-1978.09 tags: Schmidt BMI original operant conditioning cortex HOT pyramidal information antidromic date: 03-12-2019 23:35 gmt revision:11 [10] [9] [8] [7] [6] [5] [head]

PMID-101388[0] Fine control of operantly conditioned firing patterns of cortical neurons.

  • Hand-arm area of M1, 11 or 12 chronic recording electrodes, 3 monkeys.
    • But, they only used one unit at a time in the conditioning task.
  • Observed conditioning in 77% of single units and 65% of combined units (multiunits?).
  • Trained to move a handle to a position indicated by 8 annular cursor lights.
    • Cursor was updated at 50hz -- this was just a series of lights! talk about simple feedback...
    • Investigated different smoothing: too fast, FR does not stay in target; too slow, cursor acquires target too slowly.
      • My gamma function is very similar to their lowpass filter used for smoothing the firing rates.
    • 4 or 8 target random tracking task
    • Time-out of 8 seconds
    • Run of 40 trials
      • The conditioning reached a significant level of performance after 2.2 runs of 40 trials (in well-trained monkeys); typically, they did 18 runs/day (720 trials)
  • Recordings:
    • Scalar mapping of unit firing rate to cursor position.
    • Filtered 600-6kHz
    • Each accepted spike triggered a generator that produced a pulse of of constant amplitude and width -> this was fed into a lowpass filter (1.5 to 2.5 & 3.5Hz cutoff), and a gain stage, then a ADC, then (presumably) the PDP.
      • can determine if these units were in the pyramidal tract by measuring antidromic delay.
    • recorded one neuron for 108 days!!
      • Neuronal activity is still being recorded from one monkey 24 months after chronic implantation of the microelectrodes.
    • Average period in which conditioning was attempted was 3.12 days.
  • Successful conditioning was always associated with specific repeatable limb movements
    • "However, what appears to be conditioned in these experiments is a movement, and the neuron under study is correlated with that movement." YES.
    • The monkeys clearly learned to make (increasingly refined) movement to modulate the firing activity of the recorded units.
    • The monkey learned to turn off certain units with specific limb positions; the monkey used exaggerated movements for these purposes.
      • e.g. finger and shoulder movements, isometric contraction in one case.
  • Trained some monkeys or > 15 months; animals got better at the task over time.
  • PDP-12 computer.
  • Information measure: 0 bits for missed targets, 2 for a 4 target task, 3 for 8 target task; information rate = total number of bits / time to acquire targets.
    • 3.85 bits/sec peak with 4 targets, 500ms hold time
    • With this, monkeys were able to exert fine control of firing rate.
    • Damn! compare to Paninski! [1]
  • 4.29 bits/sec when the same task was performed with a manipulandum & wrist movement
  • they were able to condition 77% of individual neurons and 65% of combined units.
  • Implanted a pyramidal tract electrode in one monkey; both cells recorded at that time were pyramidal tract neurons, antidromic latencies of 1.2 - 1.3ms.
    • Failures had no relation to over movements of the monkey.
  • Fetz and Baker [2,3,4,5] found that 65% of precentral neurons could be conditioned for increased or decreased firing rates.
    • and it only took 6.5 minutes, on average, for the units to change firing rates!
  • Summarized in [1].


hide / / print
ref: -0 tags: ultrasonic BMI monkey LFP intan nordic Ozturk UCSD date: 09-30-2016 19:38 gmt revision:2 [1] [0] [head]

A Wireless 32-Channel Implantable Bidirectional Brain Machine Interface

  • Yi Su 1,2,*, Sudhamayee Routhu 2, Kee S. Moon 3, Sung Q. Lee 4, WooSub Youm 4 and Yusuf Ozturk 2,
  • Only LFP from a utah array, but solid work none-the-less.
  • 20V unipolar stimulation.
    • Through separate recording and stimulation electrodes.
  • 35mm x 10mm.
  • LFP due to limited bandwidth.
    • Less RF bw & compression that the wireless system I designed 6 years ago.
    • Reason: "Further, in order to analyze the integrative synaptic processes, LFP is the signal of interest instead of spikes, because synaptic processes cannot be captured by spike activity of a small number of neurons"
captured by spike activity of a small number of neurons.
  • Reference use of DuraGen followed by silicone elastomer.
  • Didn't cite us.

hide / / print
ref: -0 tags: noise triboelectric implant BMI date: 05-16-2014 17:28 gmt revision:1 [0] [head]

source -- Durand

hide / / print
ref: Chestek-2011.08 tags: shenoy Utah array reliability recording BMI date: 01-28-2013 20:54 gmt revision:2 [1] [0] [head]

PMID-21775782[0] Long-term stability of neural prosthetic control signals from silicon cortical arrays in rhesus macaque motor cortex (Shenoy)

  • Overall, this study suggests that action potential amplitude declines more slowly than previously supposed, and performance can be maintained over the course of multiple years when decoding from threshold-crossing events rather than isolated action potentials.
  • During most time periods, decoder performance was not well correlated with action potential amplitude (p > 0.05 for three of four arrays)
    • Perhaps we are chasing the wrong dragon?
    • Still, minimal invasiveness / more channels is useful.


[0] Chestek CA, Gilja V, Nuyujukian P, Foster JD, Fan JM, Kaufman MT, Churchland MM, Rivera-Alvidrez Z, Cunningham JP, Ryu SI, Shenoy KV, Long-term stability of neural prosthetic control signals from silicon cortical arrays in rhesus macaque motor cortex.J Neural Eng 8:4, 045005 (2011 Aug)

hide / / print
ref: -0 tags: Shenoy eye position BMI performance monitoring date: 01-25-2013 00:41 gmt revision:1 [0] [head]

PMID-18303802 Cortical neural prosthesis performance improves when eye position is monitored.

  • This proposal stems from recent discoveries that the direction of gaze influences neural activity in several areas that are commonly targeted for electrode implantation in neural prosthetics.
  • Can estimate eye position directly from neural activity & subtract it when performing BMI predictions.

hide / / print
ref: Narayanan-2005.04 tags: BMI reliability noise Laubach Yale synergy date: 01-23-2013 20:50 gmt revision:1 [0] [head]

PMID-15858046[0] Redundancy and synergy of neuronal ensembles in motor cortex.

  • Reaction time task.
  • Neurons that were the best individual predictors of task performance were not necessarily the neurons that contributed the most predictive information to an ensemble of neurons.
  • Small ensembles [of neurons] could exhibit synergistic interactions (e.g., 23 +/- 9% of ensembles with two neurons were synergistic).
  • In contrast, larger ensembles exhibited mostly redundant interactions (e.g., 99 +/- 0.1% of ensembles with eight neurons were redundant).
  • Possible interpretation: redundancy enables robustness.


[0] Narayanan NS, Kimchi EY, Laubach M, Redundancy and synergy of neuronal ensembles in motor cortex.J Neurosci 25:17, 4207-16 (2005 Apr 27)

hide / / print
ref: Ganguly-2011.05 tags: Carmena 2011 reversible cortical networks learning indirect BMI date: 01-23-2013 18:54 gmt revision:6 [5] [4] [3] [2] [1] [0] [head]

PMID-21499255[0] Reversible large-scale modification of cortical networks during neuroprosthetic control.

  • Split the group of recorded motor neurons into direct (decoded and controls the BMI) and indirect (passive) neurons.
  • Both groups showed changes in neuronal tuning / PD.
    • More PD. Is there no better metric?
  • Monkeys performed manual control before (MC1) and after (MC2) BMI training.
    • The majority of neurons reverted back to original tuning after BC; c.f. [1]
  • Monkeys were trained to rapidly switch between manual and brain control; still showed substantial changes in PD.
  • 'Near' (on same electrode as direct neurons) and 'far' neurons (different electrode) showed similar changes in PD.
    • Modulation Depth in indirect neurons was less in BC than manual control.
  • Prove (pretty well) that motor cortex neuronal spiking can be dissociated from movement.
  • Indirect neurons showed decreased modulation depth (MD) -> perhaps this is to decrease interference with direct neurons.
  • Quote "Studies of operant conditioning of single neurons found that conconditioned adjacent neurons were largely correlated with the conditioned neurons".
    • Well, also: Fetz and Baker showed that you can condition neurons recorded on the same electrode to covary or inversely vary.
  • Contrast with studies of motor learning in different force fields, where there is a dramatic memory trace.
    • Possibly this is from proprioception activating the cerebellum?

Other notes:

  • Scale bars on the waveforms are incorrect for figure 1.
  • Same monkeys as [2]


[0] Ganguly K, Dimitrov DF, Wallis JD, Carmena JM, Reversible large-scale modification of cortical networks during neuroprosthetic control.Nat Neurosci 14:5, 662-7 (2011 May)
[1] Gandolfo F, Li C, Benda BJ, Schioppa CP, Bizzi E, Cortical correlates of learning in monkeys adapting to a new dynamical environment.Proc Natl Acad Sci U S A 97:5, 2259-63 (2000 Feb 29)
[2] Ganguly K, Carmena JM, Emergence of a stable cortical map for neuroprosthetic control.PLoS Biol 7:7, e1000153 (2009 Jul)

hide / / print
ref: Hochberg-2006.07 tags: BMI Donoghue Utah probe Nature tetraplegia Hochberg 2006 date: 01-23-2013 18:49 gmt revision:4 [3] [2] [1] [0] [head]

PMID-16838014[] Neuronal ensemble control of prosthetic devices by a human with tetraplegia

  • patient was able to talk?
  • 96-channel microelectrode array implanted in arm/hand knob or right precentral gyrus.
  • around 30 units / day observed.
  • 90% of units showed significantly varied firing rates (K-S test) during imagined movements.
  • 2D control. Good pursuit tracking and center-out performance.
  • Used Wiener filter.
  • also see the technology review


hide / / print
ref: Schalk-2000.12 tags: error potential EEG wadsworth BCI 2000 BMI date: 01-23-2013 07:15 gmt revision:3 [2] [1] [0] [head]

PMID-11090763[0] EEG-based communication: presence of an error potential.

  • Idea: they trained a set of subjects to use mu/beta rhythm over central sulcus (sensorimotor) amplitude to move a cursor around the screen, and simultaneously monitored for error-related potentials to correct errors in decoding.
  • patients get 80-97% accuracy in a binary choice task.
  • look at the end of a trial to see if they 'approve' of the choice.
  • had to remove eyeblink artifacts! however, people tend to defer eyeblinks until the end of performance.
  • error = average EEG during error trials - EEG during correct trial. (a potential)
    • the error was over primary motor/ somatosensory cortex.
    • used adaptive noise cancellation to remove some of the eyeblink EMG.


[0] Schalk G, Wolpaw JR, McFarland DJ, Pfurtscheller G, EEG-based communication: presence of an error potential.Clin Neurophysiol 111:12, 2138-44 (2000 Dec)

hide / / print
ref: Chestek-2009.09 tags: BMI problems address critique spike sorting Shenoy date: 01-23-2013 02:23 gmt revision:3 [2] [1] [0] [head]

IEEE-5332822 (pdf) Neural prosthetic systems: Current problems and future directions

  • Where there is unlikely to be improvements: spike sorting and spiking models.
  • Where there are likely to be dramatic improvements: non-stationarity of recorded waveforms, limitations of a linear mappings between neural activity and movement kinematics, and the low signal to noise ratio of the neural data.
  • Compare different sorting methods: threshold, single unit, multiunit, relative to decoding.
  • Plot waveform changes over an hour -- this contrasts with earlier work (?) {1032}
  • Figure 5: there is no obvious linear transform between neural activity and the kinematic parameters.
  • Suggest that linear models need to be replaced by the literature of how primates actually make reaches.
  • Discuss that offline performance is not at all the same as online; in the latter the user can learn and adapt on the fly!


Chestek, C.A. and Cunningham, J.P. and Gilja, V. and Nuyujukian, P. and Ryu, S.I. and Shenoy, K.V. Engineering in Medicine and Biology Society, 2009. EMBC 2009. Annual International Conference of the IEEE 3369 -3375 (2009)

hide / / print
ref: -0 tags: SUA LFP BMI decoding Donoghue date: 07-24-2012 15:54 gmt revision:0 [head]

PMID-22157115 Decoding 3D reach and grasp from hybrid signals in motor and premotor cortices: spikes, multiunit activity, and local field potentials.

  • Idea: you get more information from SUA (what they call SA) activity than broadband LFPS for predicting reach direction / position for a freely moving monkey.
  • C.F. {253}

hide / / print
ref: Mehring-2003.12 tags: BMI LFP MUA SUA Mehring Vaadia date: 07-24-2012 15:54 gmt revision:3 [2] [1] [0] [head]

PMID-14634657[0]Inference of hand movements from local field potentials in monkey motor cortex

  • idea: you get equally good predictions from SUA, LFP, or MUA in decoding a 8-target center-out task.
  • c.f. {1167}


hide / / print
ref: Freire-2011.01 tags: Nicolelis BMI electrodes immune respones immunohistochemistry chronic arrays rats 2011 MEA histology date: 06-29-2012 01:20 gmt revision:5 [4] [3] [2] [1] [0] [head]

PMID-22096594[0] Comprehensive analysis of tissue preservation and recording quality from chronic multielectrode implants.

  • Says what might be expected: tungsten microelectrode arrays work, though the quality gradually declines over 6 months.
  • Histological markers correlated well with recording performance.
  • Shows persistent glial activation around electrode sites + cell body hypertropy.
    • Suggest that loss in recording quality may be due to glial encapsulation.
  • References
    • Szarowski et al 2003 {1028}
    • Ward et al 2009
  • Histology:
    • NADPH-d: nicotinamide adenine dinucleotide phosphate-diaphorase, via beta-NADP
    • CO: cytochrome oxidase, via diamnibenzidine DAB, cytochrome c and catalase.
      • both good for staining cortical layers; applied in a standard buffered solution and monitored to prevent overstaining.
  • Immunohistochemistry:
    • Activated microglia with ED-1 antibody.
    • Astrocytes labeled with glial fibrillary acid protein.
    • IEG with an antibody against EGR-1, 'a well-known marker of calcium dependent neuronal activity'
    • Neurofilament revealed using a monoclonal NF-M antibody.
    • Caspace-3 with the associated antibody
    • Details the steps for immunostaining -- wash, blocknig buffer, addition of the antibody in diluted blocking solution (skim milk) overnight, wash again, incubate in biotinylated secondary antibody, wash again, incubate in avidin-biotin-peroxidase solution.
    • Flourescent immunohistochemistry had biotynlation replaced with alexa Fluor 488-conjugated horse anti-mouse and Alexa Fluor 594-conjugated goat anti-rabbit overnight.


[0] Freire MA, Morya E, Faber J, Santos JR, Guimaraes JS, Lemos NA, Sameshima K, Pereira A, Ribeiro S, Nicolelis MA, Comprehensive analysis of tissue preservation and recording quality from chronic multielectrode implants.PLoS One 6:11, e27554 (2011)

hide / / print
ref: Rousche-1998.07 tags: BMI Utah cat Normann recording electrode MEA histology date: 06-29-2012 01:12 gmt revision:9 [8] [7] [6] [5] [4] [3] [head]

PMID-10223510 Chronic recording capability of the Utah Intracortical Electrode Array in cat sensory cortex.

  • Focus on (surprisingly) chronic recording from the utah array: they want to demonstrate that it works.
  • Platinum coating.
  • insulated with 2-3um polymide.
  • 10 cats, 12 arrays: 2 in S1, 8 in auditory ctx, 2 V1.
  • 11 electrodes connected in each array.
  • After a 6-month implant period, 60% of implanted arrays could still record 'some type of activity'.
  • They were completely targeting neuroprostheses.
    • But acknowledge that 'the presence of fibrous encapsulation and chronic astrogliosis suggests that more research is necessary before the UIEA can be uses as a cornerstone of a neuroprosthetic device for human use.
      • And yet they went through with the human trials?
  • Electrode impedance gave no hint as to the ability of a given electrode to record neural units: many electrodes with average impedance could not record neural activity.
  • Impedances generally decreased , which is not unusual (Schmidt and Bak, 1976).
    • Likely that the polymide had become permeated with water vapor to and equilibrium point. (rather than pinhole leaks or water permeation).
  • Quiet amplifiers: 2uv pk-pk.
  • No significant trend in background activity was noted over the implant durations.
  • In nearly every cat, the dura above the electrode array adhered to the bone flap, and the electrode array adhered to the dura. Therefore, when the bone flap was removed, the UIEA was concurrently explanted from the cortex.
    • Similar to Hoogerwerf and Wise 1994 {1025}
    • The explanted UIEAs typically had become encapsulated, the encapsulation was the cause of the cortical depression.
    • Only 1 did not become encapsulated in dura.
    • This encapsulation explains the gradually varying recording properties -- the electrodes were moving out of the brain.
    • "The capsule which formed around the substrate of the UIEA was usually continuous with the dura, which was enmeshed directly to the overlying skull. The encapsulated array therefore had no freedom of movement with respect to the skull, and this may have caused local trauma which reduced the possibility of recording neural activity. This relative micromovement between the fixed array and the ‘floating’ cortical tissue may also be responsible for sustaining continued growth of the encapsulation as described above."
    • Have tried putting teflon on the top of the Utah array -- did this work?
  • Two UIEAs were not found near the cortical surface -- these two arrays were totally removed from the leptomeningeal space. although originally implanted into the cortex beneath the dura, at the time of sacrafice these arrays were found above the repaired dura, and the implanted cortex showed no evicence of cortical implant.
  • Some electrodes healthy; other showed chronic inflammation.
  • General and intense inflamation in the upper layers of cortex even on their best-performing array; no guarantee that this ctx was working properly, as it is heavily compressed with fibroblasts.
  • Regarding vascluature, see {1024}.
  • Say that the largest impediment is the formation of a capsule around the implant. (Do not mention issue of infection; I guess cats have strong immune systems as well?)
  • Rather good biological discussion and conclusion. worth a re-read. "We currently recommend that the UIEA be used for acute and short-term applications."
    • Not too many follow-ups re teflon or fixing the encapsulation problem: See {1026}
      • Indeed, {1027} doesn't even cite this! Too disastrous?


hide / / print
ref: Jarosiewicz-2008.12 tags: Schwartz BMI learning perturbation date: 03-07-2012 17:11 gmt revision:2 [1] [0] [head]

PMID-19047633[0] Functional network reorganization during learning in a brain-computer interface paradigm.

  • quote: For example, the tuning functions of neurons in the motor cortex can change when monkeys adapt to perturbations that interfere with the execution (5–7) or visual feedback (8–10) of their movements. Check these refs - have to be good!
  • point out that only the BMI lets you see how the changes reflect changes in behavior.
  • BMI also allows pertubactions to target a subset of neurons. apparently, they had the same idea as me.
  • used the PV algorithm. yeck.
  • perturbed a select subset of neurons by rotating their tuning by 90deg. about the Z-axis. pre - perturb - washout series of experiments.
  • 3D BMI, center-out task, 8 targets at the corners of a cube.
  • looked for the following strategies for compensating to the perturbation:
    • re-aiming: to compensate for the deflected trajectory, aim at a rotated target.
    • re-waiting: decrease the strength of the rotated neurons.
    • re-mapping: use the new units based on their rotated tuning.
  • modulation depths for the rotated neurons did in fact decrease.
  • PD for the neurons that were perturbed rotated more than the control neurons.
  • rotated neurons contributed to error parallel to perturbation, unrotated compensated for this, and contributed to 'errors' in the opposite direction.
  • typical recording sessions of 3 hours - thus, the adaptation had to proceed quickly and only online. pre-perturb-washout each had about 8 * 20 trials.
  • interesting conjecture: "Another possibility is that these neurons solve the “credit-assignment problem” described in the artificial intelligence literature (25–26). By using a form of Hebbian learning (27), each neuron could reduce its contribution to error independently of other neurons via noise-driven synaptic updating rules (28–30). "
    • ref 25: Minsky - 1961;
    • ref 26: Cohen PR, Feigenbaum EA (1982) The Handbook of Artificial Intelligence; 27 references Hebb driectly - 1949 ;
    • ref 28: ALOPEX {695} ;
    • ref 29: PMID-1903542[1] A more biologically plausible learning rule for neural networks.
    • ref 30: PMID-17652414[2] Model of birdsong learning based on gradient estimation by dynamic perturbation of neural conductances. Fiete IR, Fee MS, Seung HS.


[0] Jarosiewicz B, Chase SM, Fraser GW, Velliste M, Kass RE, Schwartz AB, Functional network reorganization during learning in a brain-computer interface paradigm.Proc Natl Acad Sci U S A 105:49, 19486-91 (2008 Dec 9)
[1] Mazzoni P, Andersen RA, Jordan MI, A more biologically plausible learning rule for neural networks.Proc Natl Acad Sci U S A 88:10, 4433-7 (1991 May 15)
[2] Fiete IR, Fee MS, Seung HS, Model of birdsong learning based on gradient estimation by dynamic perturbation of neural conductances.J Neurophysiol 98:4, 2038-57 (2007 Oct)

hide / / print
ref: Rouse-2011.06 tags: BMI chronic DBS bidirectional stimulator Washington Medtronic ASIC translational date: 03-05-2012 23:56 gmt revision:3 [2] [1] [0] [head]

PMID-21543839[0] A chronic generalized bi-directional brain-machine interface.

  • Using a commercial neurostimulator package & battery etc.
  • "A key goal of this research prototype is to help broaden the clinical scope and acceptance of NI techniques, particularly real-time brain state detection" Good purpose! good work!
  • Augments the stimulator with 4 channels of ECoG/LFP + accelerometer + wireless telemetry.
    • Can be used to detect parkinsons state or pre-epileptiform behavior.
      • Much of this has been though of before, it just took the technology to catch up & a group to make it.
    • Chronic data is needed from humans -- animal models are often inadequate.
  • Tested in a primate for brain control of a cursor: 1D control using ECoG.
    • Good Left/right ROC, actually.
    • A large cost is simply the clinical testing; hence they piggy-back on an existing design.
    • There should be more research-industry collaborations like this.
  • impressive specs.
  • SVM classification algorithm (only consumed 10uW!) for data compression.
  • short-time Fourier transform for extracting the power over a given band. This using a modified chopper-amplification scheme. Output data has a bandwidth of less than 5Hz, which greatly reduces processing requirements.
  • Lots of processing on the BASIC chip, much like here.
  • Also see the press release


[0] Rouse AG, Stanslaski SR, Cong P, Jensen RM, Afshar P, Ullestad D, Gupta R, Molnar GF, Moran DW, Denison TJ, A chronic generalized bi-directional brain-machine interface.J Neural Eng 8:3, 036018 (2011 Jun)

hide / / print
ref: Benabid-2011.01 tags: DBS BMI Benabid date: 02-16-2012 17:48 gmt revision:3 [2] [1] [0] [head]

PMID-21867795[0] Deep brain stimulation: BCI at large, where are we going to?

  • Everybody on the bandwagon! Talks about their efforts to make a BCI using DBS techniques / electrodes.
  • Language is a bit telegraphic, perhaps because it's translated from the French.
  • Tested in rats using an 1D ECoG BMI.
  • Posits that DBS is just a particularly simple form of BMI.


[0] Benabid AL, Costecalde T, Torres N, Moro C, Aksenova T, Eliseyev A, Charvet G, Sauter F, Ratel D, Mestais C, Pollak P, Chabardes S, Deep brain stimulation: BCI at large, where are we going to?Prog Brain Res 194no Issue 71-82 (2011)

hide / / print
ref: Leach-2010.02 tags: BMI challenges histology biocompatibility review date: 01-16-2012 18:22 gmt revision:4 [3] [2] [1] [0] [head]

PMID-20161810[0] Bridging the Divide between Neuroprosthetic Design, Tissue Engineering and Neurobiology

  • Neuroprosthetic device technology has seen major advances in recent years but the full potential of these devices remains unrealized due to outstanding challenges, such as the ability to record consistently over long periods of time.
  • Discuss promising new treatments based on developmental and cancer biology (?)
  • Suggest controlled drug release as the tissue is healing. Makes sense.


[0] Leach JB, Achyuta AK, Murthy SK, Bridging the Divide between Neuroprosthetic Design, Tissue Engineering and Neurobiology.Front Neuroeng 2no Issue 18 (2010 Feb 8)

hide / / print
ref: Ganguly-2009.07 tags: Ganguly Carmena 2009 stable neuroprosthetic BMI control learning kinarm date: 01-14-2012 21:07 gmt revision:4 [3] [2] [1] [0] [head]

PMID-19621062 Emergence of a stable cortical map for neuroprosthetic control.

  • Question: Are the neuronal adaptations evident in BMI control stable and stored like with skilled motor learning?
    • There is mixed evidence for stationary neuron -> behavior maps in motor cortex.
      • It remains unclear if the tuning relationship for M1 neurons are stable across time; if they are not stable, rather advanced adaptive algorithms will be required.
  • A stable representation did occur.
    • Small perturbations to the size of the neuronal ensemble or to the decoder could disrupt function.
    • Compare with {291} -- opposite result?
    • A second map could be learned after primary map was consolidated.
  • Used a Kinarm + Plexon, as usual.
    • Regressed linear decoder (Wiener filter) to shoulder and elbow angle.
  • Assessed waveform stability with PCA (+ amplitude) and ISI distribution (KS test).
  • Learning occurred over the course of 19 days; after about 8 days performance reached an asymptote.
    • Brain control trajectory to target became stereotyped over the course of training.
      • Stereotyped and curved -- they propose a balance of time to reach target and effort to enforce certain firing rate profiles.
    • Performance was good even at the beginning of a day -- hence motor maps could be recalled.
  • By analyzing neuron firing wrt idealized movement to target, the relationship between neuron & movement proved to be stable.
  • Tested to see if all neurons were required for accurate control by generating an online neuron dropping curve, in which a random # of units were omitted from the decoder.
    • Removal of 3 neurons (of 10 - 15) resulted in > 50% drop in accuracy.
  • Tried a shuffled decoder as well: this too could be learned in 3-8 days.
    • Shuffling was applied by permuting the neurons-to-lags mapping. Eg. the timecourse of the lags was not changed.
  • Also tried retraining the decoder (using manual control on a new day) -- performance dropped, then rapidly recovered when the original fixed decoder was reinstated.
    • This suggests that small but significant changes in the model weights (they do not analyze what) are sufficient for preventing an established cortical map from being transformed to a reliable control signal.
  • A fair bit of effort was put into making & correcting tuning curves, which is problematic as these are mostly determined by the decoder
    • Better idea would be to analyze the variance / noise properties wrt cursor trajectory?
  • Performance was about the same for smaller (10-15) and larger (41) unit ensembles.

hide / / print
ref: Parikh-2009.04 tags: BMI rats cortex layer depth date: 01-10-2012 01:09 gmt revision:2 [1] [0] [head]

PMID-19255460[0] Lower layers in the motor cortex are more effective targets for penetrating microelectrodes in cortical prostheses.

  • Aggregate analysis (633 neurons) and best session analysis (75 neurons) indicated that units in the lower layers (layers 5, 6) are more likely to encode direction information when compared to units in the upper layers (layers 2, 3) (p< 0.05).
  • DUH. Have we forgotten all anatomy?


[0] Parikh H, Marzullo TC, Kipke DR, Lower layers in the motor cortex are more effective targets for penetrating microelectrodes in cortical prostheses.J Neural Eng 6:2, 026004 (2009 Apr)

hide / / print
ref: Santhanam-2006.07 tags: Shenoy BMI BCI trials date: 01-08-2012 23:37 gmt revision:4 [3] [2] [1] [0] [head]

PMID-16838020[0] A high-performance brain-computer interface

  • the speed and accuracy with which keys can be selected using BCIs is still far lower than for systems relying on eye movements.
    • What is the eye-movement rate?
  • implanted in PMD. 96 electrodes (utah array).
  • used an instructed-delay task. figure 1
    • monkey had to move to target when center target disappeared. peripheral target appeared several seconds prior.
  • actually had the monkey reach to targets; if correct, monkey was immediately rewarded.
    • real movement trials were interspersed to keep the monkey engaged.
  • decoding model: assume that the spike counts come from a poisson or gaussian distribution. Apply ML decoding.
    • poisson better than gaussian.
  • up to 6.5 bits per second, or approximately 15 words per minute, with 96 electrodes.
    • Peak of continuous control = 1.6 bits per second.
  • ITRC = information transfer rate capacity. this metric is proportional to the single trial accuracy / trial length (sorta, see ref 23 - Blahut-Arimoto algorithm)
  • most of their neurons seem to be responsive to actual movements (que supressa!)
  • maximum bandwidth with a trial length of 250ms.
    • lots of other good information-theoretic analysis.
  • PMID-12657892[1] Neural prosthetic control signals from plan activity. -- the preceding Neuroreport simulation study.
    • performance to exceed 90% with as few as 40 neurons.
    • maximum likelihood decoders controlling a FSM.


hide / / print
ref: Musallam-2004.07 tags: cognitive BMI Musallam Andersen PRR MIP date: 01-08-2012 23:13 gmt revision:5 [4] [3] [2] [1] [0] [head]

PMID-15247483[0] Cognitive control signals for Neural Prosthetics

  • decode intended target from 200 to 1100ms of memory period (reward on correct, etc).
  • got good success rates with relatively few neurons (like 8 for 8 targets) -- yet decode rates were not that good, not at all as good as Fetz or Schmidt.
  • used pareital reach region (PRR), a subsection of posterior partietal cortex PPC, which represents the goals of the reach in visual coordinates. In the experiment, the implanted in media intrapareital (MIP)
    • in encodes the intended goal rather than the trajectory to achieve that goal.
    • PMd also seems to encode planning activity, though less is known about that.
  • used an adaptive database to map neuronal activity to targets; eventually, the database contained only (correct) brain-control trials.
  • neuronal responses were recorded from parietal reach region (PRR) with 64 microwire electrodes in 4 monkeys, plus 32 microwire electrodes in PMd
  • monkeys were tained to fixate on the center of the screen dring the task, though free fixation was also tested and seemed to work ok.
  • monkeys had to press cue, fixate, observe target location, wait ~2 sec, and move to the (remembered) target location when cue disappeared.
  • they use a static or continually updated 'database' for predicting which of four targets the monkey wants to go to during the instructed delay task.
  • able to predict with moderate accuracy the expected value of the target as well as its (discrete) position.
  • predictions were made during the delay period while there was no motor movement.
  • predictions worked equally well for updated and static databases.
  • monkeys were able to increase their performance on the BMI trials over the course of training.
  • reward type or size modulated the tuning of BMI neurons in the ecpected way, though aversive stimuli did not increase the tuning - suggesting that the tuning is not a function of attention (maybe).
  • the database consisted of 900ms of spike recordings starting 200ms after cue for 30 reach trials for each target. spike trains were projected onto Haar wavelets (sorta like a binary tree), and the filter coefficients were used to describe P(r), the probability of response, and P(r|s), probability of response given the target. then they used bayes rule (P(r) and P(r|s) were approximated with histograms, i think) to find P(s|r) - a discrete function - which it is easy to find the maximum of.
  • adding more trials offline improved the decode performance.
  • supporting online material.

PMID-15491902 Cognitive neural prosthetics

  • LFPs are easier to record and may last longer (but they are not as 'sexy').
  • suggest future electrodes will move automatically, peizo-drive perhaps.
  • PRR receives direct visual projections & codes for reaches in visual coordinates relative to the current direction of gaze.
  • PRR can hold the plan for a movement in short-term memory.
  • 16 neurons peak..?
  • In area LIP of PPC Platt and Glimcher PMID-10421364 found cells that code the expected value of rewards.
    • 20Hz beta-band oscillation indicated the behavioral state of the animal. While planning for a saccade it slowly increased, whereas at the time of movement in dramatically increased in amplitude.
    • LFP was better than spikes for a state decode.


hide / / print
ref: Bossetti-2004.06 tags: Bossetti wolf Carmena Nicolelis latency wireless BMI recording date: 01-08-2012 21:16 gmt revision:2 [1] [0] [head]

IEEE-1300783 (pdf) Transmission latencies in a telemetry-linked brain-machine interface

  • quote: "examines the relationships between the ratio of output to average input bandwidth of an implanted device and transmission latency and required queue depth".
  • can use to explain why I decided on the fixed-bandwidth method. must measure the latency on my system .. how?
  • firing bursts results in high latencies in a variable-bandwidth queued system.
  • Tested in 32-neuron ensemble.
  • require output bandwidth / input bandwidth to be at least 4 to get sub-10ms max latency.


Bossetti, C.A. and Carmena, J.M. and Nicolelis, M.A.L. and Wolf, P.D. Transmission latencies in a telemetry-linked brain-machine interface Biomedical Engineering, IEEE Transactions on 51 6 919 -924 (2004.06)

hide / / print
ref: Carmena-2003.11 tags: Carmena nicolelis BMI learning 2003 date: 01-08-2012 18:53 gmt revision:5 [4] [3] [2] [1] [0] [head]

PMID-14624244[0] Learning to control a brain-machine interface for reaching and grasping by primates.

  • strong focus on learning & reorganization.
  • Jose's first main paper.
  • focuses on two engineering / scientific questions: what signal to use, and how much of it, and from where.
    • As for where, of course we suggest that the representation is distributed.
  • Quality of predictions: gripping force > hand velocity > hand position.
  • Showed silent EMGs during BMI control.
  • Put a robot in the feedback path; this ammounted for some nonlinearities + 60-90ms delay.
  • Predictions follow anatomical expectation:
    • M1 (33-56 cells) predicts 73% variance for hand pos, 66% velocity, 83% for gripping force .
    • SMA (16-19 cells) 51% position, 51% velocity, 19% gripping force.
    • They need a table for this shiz.
  • Relatively high-quality predictions. (When I initially looked at the data, I was frustrated with the noise!)
  • Learning was associated with increased contribution of single units.
    • appeared to be more 'learning' in SMA.
    • Training on a position model seemed to increase the ctx representation of hand position.
  • changes between pole control and brain control:
    • 68% of of sampled neurons showed reduced tuning in BCWOH
    • 14% no change
    • 18% enhanced tuning.
  • Directional tuning curves clustered in a band during brain control -- neurons clustering around the first PC?
    • All cortical areas tested showed increases in correlated firing -- arousal?
    • this puts some movements into the nullspace of the Wiener matrix. Or does it? should have had the monkey make stereotyped movements to dissociate movement directions.
  • Knocks {334} in that:
    • preferred directions were derived not from actual movements, but from firing rates during target appearance time windows.
    • tuning strength could have increased simple because the movements became straighter with practice.
  • From Fetz, {329}: Interestingly, the conversion parameters obtained for one set of trials provided increasingly poor predictions of future responses, indicating a source of drift over tens of minutes in the open-loop condition. This problem was alleviated when the monkeys observed the consequences of their neural activity in ‘real time’ and could optimize cell activity to achieve the desired goal under ‘closed-loop’ conditions.


hide / / print
ref: Moritz-2008.12 tags: FES BMI Fetz Moritz Perlmutter spinal cord date: 01-08-2012 05:18 gmt revision:1 [0] [head]

PMID-18923392[0] Direct control of paralysed muscles by cortical neurons.

  • FES BMI: route signals around a broken spinal cord.
  • Found that "neurons could control functional stimulation equally well regardless of any prior association to movement". interesting. consistent with previous work. Wonder if I can duplicate this result.
  • Another relatively straightforward (?) paper where most of the difficulty is technology (!!). I mean, what new knowledge was needed to do this? Compare this with the technology that was needed. One of these was very challenging. now, as it come in for my stuff: what does technology let you do? Have to motivate.


[0] Moritz CT, Perlmutter SI, Fetz EE, Direct control of paralysed muscles by cortical neurons.Nature 456:7222, 639-42 (2008 Dec 4)

hide / / print
ref: Taylor-2002.06 tags: Taylor Schwartz 3D BMI coadaptive date: 01-08-2012 04:29 gmt revision:7 [6] [5] [4] [3] [2] [1] [head]

PMID-12052948[0] Direct Cortical Control of 3D Neuroprosthetic Devices

  • actually not a bad paper... reasonable and short. they adapted the target size to maintain a 70% hit rate, and one monkey was able to floor this (reach and stay at the minimum)
  • coadaptive algorithm removed noise units based on (effectively) cross-validation.
    • both arms were restrained during performance & co-adaptation. Monkeys initially strained to move the cursor, but eventually relaxed.
  • Changes from hand control to brain control random but apparently somewhat consistent between days.
  • continually increasing performance in brain-control for both monkeys, arguably due to the presence of feedback and learning. They emphasize the difference between open-loop (Wessberg) and closed-loop control. (42 ± 5% versus 12 ± 5% of targets hit)
    • still, the percentage of correct trials is low - ~50% for the 8 target 3D task.
    • monkeys improved target hit rate by 7% from the first to the third block of 8 closed-loop movements each day.
  • claim that they were able to record some units for up to 2 months ?? ! In their other monkey, with teflon/polymide coated stainless electrodes, the neural recordings changed nearly every day, and eventually went away.
  • quote: Cell-tuning functions obtained during normal arm movements were not good predictors of intended movement once both arms were restrained. interesting.
  • coadaptive algorithm:
    • Raw PV yielded poor predictions.
    • first, effectively z-score the firing rate of each neuron.
    • junk / hash neurons were not removed.
    • Two different weights per neuron per axis (hence 6 weights altogether), one if firing rate was above the mean value, another if it was below. corrected for resulting drift. Sum (neuronal firing rates * weights) controlled velocity on each of the axes. (Hence, it is not surprising that the brain-control tuning was significantly different from the hand control - the output model is vastly different).
    • restarted the coadaptive algorithm every day?
    • coadaptive algorithm appears to be something like stochastic gradient descent with a step-size that decreases with increasing performance.
      • From her Case-western website, Dawn Taylor still seems to be on the coadaptive kick. Seems like it's bad to get stuck on one idea all your life ... though perhaps that is the best way to complete something.
    • Their movies in supplementary materials look rather good, better than most of the stuff that we have done. She did not quantify SNR or correlation coefficient.


hide / / print
ref: Serruya-2003.03 tags: BMI Serruya Donoghue date: 01-08-2012 03:31 gmt revision:5 [4] [3] [2] [1] [0] [head]

PMID-12647229[0] Robustness of neuroprosthetic decoding algorithms

  • chronic intracortical arrays
  • step tracking and slow continuous tracking tasks
  • compared two models:
    • ML model to predict reaching directions
    • linear model to predict hand trajectory
  • Less than 1 min of data for the discrete task (8 to 13 neurons) and less than 3 min (8 to 18 neurons) for the continuous task were required to build optimal models
    • however, their definition of 'optimal' is discounted for the cost of training the model.
    • increasing the time between training and applying the model did not significantly impact the efficacy of the predictor.
      • linear predictors trained on one day & tested on another were not differentn in error characteristics from linear filters trained on the same day (!!!)
  • point out that more neurons most likely means a longer model training time
  • interesting facts: 600,000 patients in the US suffer each year from motor impairment de to spinal-cord and brainstem trauma (that's a lot; I kinda don't trust that number!)


hide / / print
ref: Wessberg-2000.11 tags: wessberg nicolelis BMI wiener date: 01-08-2012 02:53 gmt revision:4 [3] [2] [1] [0] [head]

PMID-11099043[0] Real-time prediction of hand trajectory by ensembles of cortical neurons in primates

  • both linear and nonlinear methods
  • 3d robotic control through the internet gee-whiz!
  • 2 owl monkeys.
  • Showed neuron-dropping curves.
    • Analysis revealed fewer PMd neurons would be required to achieve 90% accuracy (480 PMD, approximately 660 M1, 1200 ipsilateral).
  • Used non-overtrained food picking behavior.


hide / / print
ref: Kennedy-1998.06 tags: Kennedy cone electrode ALS BMI date: 01-08-2012 00:35 gmt revision:4 [3] [2] [1] [0] [head]

PMID-9665587[0] Restoration of neural output from a paralyzed patient by a direct brain connection.

  • (abstract) Patients with severe paralysis of limbs, face and vocal apparatus may be intelligent and aware and yet, tragically, unable to communicate. We describe a communication link for such a 'locked-in' patient with amyotrophic lateral sclerosis. We recorded action potentials in her brain over several months by means of an electrode that induces growth of myelinated fibers into its recording tip. She was able to control the neural signals in an on/off fashion. This result is an important step towards providing such patients with direct control of their environment by interfacing with a computer. Additionally, it indicates that restoration of paralyzed muscles may be possible by using the signals to control muscle stimulators.
  • Repairing the PNS is hard. I wonder if a more logical, Gates-ian topic of socially worthwhile work would be to target more people. For example: curing cancer. There is a bit of vanity and scifi chasing in BMI studies mixed up with all the curiosity (science) and worthy causes.
    • That said, plenty of people are doing shit with their lives (in terms of social worth). All that realyl matters is passion. Kenny seems to have passion for his particular recording / BMI idea, which is good!


[0] Kennedy PR, Bakay RA, Restoration of neural output from a paralyzed patient by a direct brain connection.Neuroreport 9:8, 1707-11 (1998 Jun 1)

hide / / print
ref: Kennedy-2000.06 tags: BMI Kennedy IEEE 1999 Bakay neurotrophic glass electrode date: 01-08-2012 00:30 gmt revision:7 [6] [5] [4] [3] [2] [1] [head]

PMID-10896186[] Direct control of a computer from the human central nervous system

  • 1999 - eight years ago!
  • Se also {1020}
  • 3 patients, one success. (one died of ALS :-( JR had a brainstem stroke.
    • Has disconjugate eye movements with nystagmus.
    • These patients are sick - ulcers, peripheral neuropathy.
  • invasive alternative to externally applied BCI.
  • one glass electrode
  • patient can type & produce synthesized speech.
  • cursor wraps off the right of the screen; only positive deflections matter. another signal or dwell used to select letters.
  • Used surface EMG to dissociate neural activity from muscle contraction, as in previous works.
  • electrode used in this study: [1] "The cone electrode: ultrastructural studies following long-term recording in rat and monkey cortex."


hide / / print
ref: Wyler-1980.08 tags: Wyler operant conditioning fast slow pyramidal tract neurons BMI date: 01-07-2012 22:09 gmt revision:3 [2] [1] [0] [head]

PMID-7409057[0] Operant control of precentral neurons: comparison of fast and slow pyramidal tract neurons.

  • Slow PTN (neurons with antidromic latency > 2ms) are pratically all well controlled in his operant-conditioning task;
  • Fast (< 2ms, mean 1.2ms latency) have a more highly variable firing rates and ISIs.
  • "[I]t appears that the majority of error from fast PT cells was generated by ISIS less than 30 ms, whereas the majority of error for slow PT cells was represented in ISIS greater than 60 ms."
    • Ok, trivial observation, but still interesting.


[0] Wyler AR, Burchiel KJ, Robbins CA, Operant control of precentral neurons: comparison of fast and slow pyramidal tract neurons.Exp Neurol 69:2, 430-3 (1980 Aug)

hide / / print
ref: Fetz-1973.03 tags: operant conditioning Fetz Baker learning BMI date: 01-07-2012 19:34 gmt revision:2 [1] [0] [head]

PMID-4196269[0] Operantly conditioned patterns on precentral unit activity and correlated responses in adjacent cells and contralateral muscles

  • Looked at an operant task through the opposite direction: as a means for looking at reaction time, and muscle responses to trained bursts of activity.
  • recorded from precentral gyrus cells in leg and arm representation.
    • isonel coated tungsten microwires, with great apparent waveform records.
  • also recorded EMG, nylon-insuldated stainless-steel wire, led subcutaneuosly to the head connector.
  • references an even older study concerning the operant conditioning of neural activity in rats by Olds.
  • really simple technology - RC filter to estimate the rate; reward high rate; resets on reward.
    • the evoked operant bursts are undoubtably due to training.
  • looks like it was easy for the monkeys to increase the firing rate of their cortical cells (of course, I'm just skimming the article..)
  • 233 precentral units.
    • which they did some preliminary somatotopic mapping of.
  • neighboring cells mirrored the firing rate changes (logical as they share the local circuitry)
  • in a few sessions the operant bursts were not associated with movements.
  • Could individually condition cells when they happened to record 2 units on the same electrode.


hide / / print
ref: Fetz-1969.02 tags: BMI original Fetz operant conditioning date: 01-07-2012 19:04 gmt revision:7 [6] [5] [4] [3] [2] [1] [head]

PMID-4974291[0] Operant conditioning of cortical unit activity

  • (Abstract) The activity of single neurons in precentral cortex of unanesthetized monkeys (Macaca mulatta) was conditioned by reinforcing high rates of neuronal discharge with delivery of a food pellet. Auditory or visual feedback of unit firing rates was usually provided in addition to food reinforcement. After several training sessions, monkeys could increase the activity of newly isolated cells by 50 to 500 percent above rates before reinforcement.
  • Used 'classical' single unit recording.
  • Trepination 5mm circle over hand area.
  • feedback: click for each AP.
  • reinforced on neuron per day.
  • trained neural activity often bursts, usually involved movement such as flexion of the lebow or rotation of the wrist.
  • controlled for sensory positive-feedback loop by performing extinction trials & looking for PETH response to click.
  • I gotta get one of these pellet feeders. monkeys will likely be more motivated, especially if I titrate how frequently they get the food.
  • images/303_1.pdf

PMID-5000088[1] Operant conditioning of specific patterns of neural and muscular activity.

In awake monkeys we recorded activity of single "motor" cortex cells, four contralateral arm muscles, and elbow position, while operantly reinforcing several patterns of motor activity. With the monkey's arm held semiprone in a cast hinged at the elbow, we reinforced active elbow movements and tested cell responses to passive elbow movements. With the cast immobilized we reinforced isometric contraction of each of the four muscles in isolation, and bursts of cortical cell activity with and without simultaneous suppression of muscle activity. Correlations between a precentral cell and specific arm muscles consistently appeared under several behavioral conditions, but could be dissociated by reinforcing cell activity and muscle suppression.

PMID-4624487[2] Operant conditioning of isolated activity in specific muscles and precentral cells

Recorded precentral units in monkeys, trained to contract 4 arm muscles in isolation, under various conditions: passive movements and cutaneous stimulation, active movements and isometric contractions. Some Ss were also reinforced for activity of cortical cells, with no contingency in muscle activity and with simultaneous suppression of all muscular activity. It is concluded that temporal correlations between activity of precentral cells and some other component of the motor response, e.g., muscle activity, force, or position, may depend as strongly on the specific response pattern which is reinforced as on any underlying physiological connection.


hide / / print
ref: Olds-1967.01 tags: Olds 1967 limbic system operant conditioning recording rats electrophysiology BMI date: 01-06-2012 03:59 gmt revision:2 [1] [0] [head]

PMID-6077726[0] The limbic system and behavioral reinforcement

  • Can't seem to find Olds 1965, as was a conference proceeding .. this will have to do, despite the lack of figures. images/966_1.pdf
  • First reference I can find of chronic (several weeks) (4-9 microelectrodes, single) recording from the rat.
  • Basically modern methods: commutator + solid state preamplifiers mounted to a counterbalanced slack-relieving arm.
    • If unit responses were observed in recordings from a given probe a week after surgery they were usually recordable indefinitely. 44 years later ...
  • Used a primitive but effective analog spike discriminator based on:
    • minimum amplitude
    • maximum amplitude
    • minimum fall time
    • maximum fall time.
  • Also had a head movement artifact detector, which blanked the recordings (stopped the paper roll) for 2 sec.
  • Reinforced on 'bursting', threshold sufficiently high that it only occurred once every 5-15 minutes.
  • Food reinforcement or 1/4 second train of brain stimulation (30ua, 60Hz, sine, in hypothalamus).
  • Reinforcement was conditioned on an 'acquisition' signal, which is visual (?) Bursting is rewarded for 2 minutes, ignored for 8 minutes.
  • Also recorded control neurons.
  • (they were looking at these things as though anew!) "The most striking aspect of the records so formed [on sheets of paper] was that all discriminators at one time or another exhibited rate changes that had the appearance of waves with a period of 10 to 20 minutes. Waves between units in the same animal were to some degree synchronized." Then describes a ramp ..
  • Longer term variations: FR would vary by a factor of 2-5 over a period of several hours.
    • This would make negatively correlated neurons (on a short time scale) appear positively correlated over long time scales (have to fix this in the BMI!)
  • As this was a conditional reinforcement task, they unexpectedly found that the acquisition periods were systematically different than extinction periods
    • More like pavlovian conditioning, esp in the hippocampus, where a conditioned response was also reflected on a control neuron.
    • Even when the light was lit throughout the acquisition period was replaced by a bell at the beginning of the acq. period, there was still a sustained change in FR.
      • Then during the extinction period: it appeared from the record of responses that a definite operant behavior was tried several times and then stopped altogether."
  • In the pontine nucleus (relay from M1 to cerebellum, v. roughly), judging from the control responses, all were conditioned.
    • Pontine responses seem to correspond with movement of the eyes or head that did not set off the movement detector/blanker.
  • Saw brief and very fast bursts during the extinction periods of the kind that Evarts found to characterize pyramical neurons during sleep.
  • When units shifted from food reward to ICS reward, units became undiffarentiated, and within a day they would be reconditioned.
  • Also tried paralyzing the animal to see if it could still generate operant responses; the animal died, results inconclusive.
  • Flood lights made it hard for the rats to produce the operant behavior.


[0] Olds J, The limbic system and behavioral reinforcement.Prog Brain Res 27no Issue 144-64 (1967)

hide / / print
ref: notes-0 tags: motor control BMI M1 date: 01-06-2012 03:11 gmt revision:13 [12] [11] [10] [9] [8] [7] [head]

with: {277}

  • Correlations have been described between neuronal activity and the static and dynamic forces and torques generated across single joints
  • or by the whole arm
  • or by precision pinch [14,15,16]
  • or there are strong correlations to muscle activity [17,18,19,20,21,22,23,24,25]
  • or there is strong correlations to kinematic parameters
  • these kinematic parameters are dependent on location in the external workspace [10][28,29]
  • kinematic tuning can be subserved by training! [30]
  • distance to target representation [31]


hide / / print
ref: Zacksenhouse-2007.07 tags: Zacksenhouse 2007 Odoherty Nicolelis cortical adaptation BMI date: 01-06-2012 03:10 gmt revision:3 [2] [1] [0] [head]

PMID-17637835[0] Cortical modulations increase in early sessions with brain-machine interface.

  • "we demonstrate that the modulations of the firing-rates of cortical neurons increased abruptly after the monkeys started operating the BMI"
    • My hypothesis: is this like LMAN? Injection of noise for the purpose of exploration?
    • Their hypothesis: we are listening to the noise or effect of increased processing / congnitive load.
    • Alternative: decreased feedback / scrabled feedback makes the individual control signals themselves less controlled.
  • Describes spikes as inhomogeneous poisson processes, and breaks things down thusly.
  • Also develop a parametric model of neuronal firing based on tuning to movement, including velocity and acceleration.
  • Fano factor of recorded neurons increased during BCWH & BCWOH.
  • Percent overall modulation (POM) higher in brain control. That is, the variance explained not by the inhomogeneous poisson process, but rather by firing rate variations.
    • "[T]he ensemble-POM increased mainly due to an increase in the variance of the spike-count, which was not matched by the change in the mean spike-count."
  • Figure 6 is pretty convincing, actually.
  • PVM (percent velocity modulation) correlates strongly with POM, but with a fractional slope, indicating that veolocity tuning accounts for only a fraction of the variance.
    • "Since the increase in POM was not matched by increasing PVM or PKM, the higher neuronal rate modulations observed during brain control cannot be explained only by increased modulations due to the kinematics of the movement."


[0] Zacksenhouse M, Lebedev MA, Carmena JM, O'Doherty JE, Henriquez C, Nicolelis MA, Cortical modulations increase in early sessions with brain-machine interface.PLoS One 2:7, e619 (2007 Jul 18)

hide / / print
ref: -2008 tags: OCZ NIA teardown autopsy BMI BCI date: 01-06-2012 03:09 gmt revision:19 [18] [17] [16] [15] [14] [13] [head]

Recently we bought a OCZ NIA device for our lab. Having designed similar hardware myself, I simply *had to* take the thing apart to inspect it, as others have done -- see Joe Pit's teardown (with schematic!!). Of course, I graciously let the others try it for a few hours (it doesn't work all that well) before taking the anodized, extruded, surface- ground aluminum case apart. Below is the top side of the 4-layer circuit board inside the case, as well as a key to indicate the function of the labeled devices. (some of the labels are hard to read due to the clutter of the silkscreen on the board; sorry).

  • A - Input connector. Center channel is isolated ground; outside two channels are the signal. They had to make this custom so people couldn't plug it into other (possibly dangerous) stuff.
  • B - Input current limiting resistors, in series with signal, 4.02K
  • C - Dual capacitor from input channels to shared ground (I think; the cap has 4 contacts, 2 at the end, 2 in the middle; I assume they use this package to get very accurately matched capacitance so as not to hurt the CMRR of the instrumentation amplifier).
  • D - Gain-setting resistor, 1.00K. Sets the instrumentation amplifier gain to 50 (I think).
    • I do not know what devices were intended for the 1206 footprints above and below this resistor...
  • E - Instrumentation amplifier, Analog Devices logo, AD8220 by my guess, A-grade. Measures the difference in voltage between the two input channels (left and right electrodes on the headband).
  • F - 47 ohm resistors & capacitors to filter the power supply to the instrumentation amplifier.
  • H - Opamp, Texas Instruments OPA348A. Looks like it is used as feedback to the instrumentation amplifier reference pin to effect highpass operation (?).
  • I - Quad opamp, TI OPA4348A. Used to filter the signal; I did not go through the filter topology, but they might have copied it off the AD8220 datasheet ;)
  • J - Stereo ADC, Texas Instruments (Burr-Brown logo, TI bought BB) PCM1803A. Only one channel is used. 24 bits, 96khz max sampling rate; device in master mode (Mode1 = 0V, Mode0 = 3.3v); Fs = SCLK/512 -> sampling rate = 3.90625 KHz.
  • K - Three channel digital isolator, Analog Devices ADUM1300. Transmits the ADC's DOUT, BCK, and LRCLK signals to the USB (non-isolated) side.
  • L - Two-channel optical (?) isolator; unknown type; used to drive the ADC's SCLK and some other signal ?
    • from Joe Pits: "Yeah, optical isolator with logic gates for high speed I guess (HCPL2631S). I'm also not sure what the second signal does, it goes to U4 (JSR marking). I suspect it could be a switch which adds C14 + R17 in the feedback loop of U2C (see the schematic). But I don't know what the reason for this is."
  • M - Isolated supply daughterboard, Texas Instruments logo, very simple design: driver is 2 BJTs (which get hot!) in push-pull topology; bases are driven by windings on the toroidal transformer; transformer center tap seems to go to USB VCC. Output is +-5V.
  • N - +3.9V, +3.3, and -3.9V power supply circuitry. I cannot identify the SOT-23-5's and SC-70's here.
  • O - PIC18F2455, with USB 2.0 (obviously!) SOIC-28 package.
    • device comes up as (on my Linux box, Debian Lenny, kernel 2.6.24):
      • usb 4-1: new full speed USB device using uhci_hcd and address 8
      • hiddev96hidraw1: USB HID v1.10 Device [Brain Actuated Technologies Neural Impulse Actuator Prototype 1.0] on usb-0000:00:1a.1-1
    • I'll put up a usbmon trace later, maybe.
  • P - Transistors for driving the tricolor LEDS on the bottom of the board.
  • Q - 16.0000 MHz crystal. Needed for correct USB timing; clocks the PIC at 48Mhz.
  • R - USB type B connector. Note the ferrites to the left. (I though they were fuses, but I accidentally shorted Vdd to ground while probing the programming connector, and these let out a little smoke rather than blowing completely. Had they been fuses, they would be open circuit now. This is consistent with Joe Pit's analysis.)
  • S - 74HCT595A 8-bit shift registers, to convert the serial data into parallel data for the PIC to read in. 3 devices = 24 bits in total.
    • Note that the 74HCT595A has a output enable, which permits the PIC to read the 3 bytes of the sample sequentially. Otherwise, as Stefan Jung (via the openeeg-list) points out, the PIC would not have enough data pins (28 pins vs. 24 bits)!
  • T - 74HCT393, Texas Instruments logo, Dual 4-bit binary ripple counter. Used to drive the ADC with a 2Mhz clock, which puts the sampling rate at (as before) 3.90625 KHz.
  • U - Programming connector. That's right, a programming connector! Looks to be the same as a PIC ICSP connector (pointed out on hack a day)
    • So far as I can tell:
      • Pin 1 = +5V, PIC pin 1, (through 100 ohm resistor), Vpp (?)
      • Pin 2 = PIC pin 20 , Vdd
      • Pin 3 = PIC pin 19 , Vss
      • Pin 4 = PIC pin 28 (through 100 ohm resistor), PGD
      • Pin 5 = PIC pin 27 (through 100 ohm resistor), PGC
    • I do not know if the device can be reprogrammed, though it looks that way.
    • from here - bootloader (to address 0x07ff) can be read, but everything above that is read-protected.
Bottom of board, showing the (very bright!) tricolor LEDs


hide / / print
ref: notes-2000.09 tags: BMI recording technology Chapin Nicolelis battery Wolf date: 01-06-2012 03:09 gmt revision:4 [3] [2] [1] [0] [head]

from the book "Neural Prostheses for Restoration of Sensory and Motor Function" edited by John Chapin and Karen Moxon.

Phillip Kennedy's one-channel neurotrophic glass electrode BMI (axons apparently grew into the electrode, and he recorded from them)

Pat Wolf on neural amplification / telemetry technology

battery technology for powering the neural telemetry

hide / / print
ref: bookmark-2007.08 tags: donoghue cyberkinetics BMI braingate date: 01-06-2012 03:09 gmt revision:3 [2] [1] [0] [head]

images/425_1.pdf August 2007

  • provides more extensive details on the braingate system.
  • including, their automatic impedance tester (5mv, 10pa)
  • and the automatic spike sorter.
  • the different tests that were required, such as accelerated aging in 50-70 deg C saline baths
  • the long path to market - $30 - $40 million more (of course, they have since abandoned the product).

hide / / print
ref: web-1972 tags: BCI BMI silly groovy date: 01-06-2012 03:07 gmt revision:2 [1] [0] [head]


  • since 1972 - groovy!
  • how does this company stay afloat?
  • looks like they have products & software for Mac & PC

hide / / print
ref: Fei-2011.05 tags: flash FPGA neural decoder BMI IGLOO f date: 01-06-2012 00:20 gmt revision:2 [1] [0] [head]

IEEE-5946801 (pdf) A low-power implantable neuroprocessor on nano-FPGA for Brain Machine interface applications

  • 5mW for 32 channels, 1.2V core voltage.
  • RLE using thresholding / transmission of DWT coefficients.
  • 5mm x 5mm.


Fei Zhang and Aghagolzadeh, M. and Oweiss, K. Acoustics, Speech and Signal Processing (ICASSP), 2011 IEEE International Conference on 1593 -1596 (2011)

hide / / print
ref: Dethier-2011.28 tags: BMI decoder spiking neural network Kalman date: 01-06-2012 00:20 gmt revision:1 [0] [head]

IEEE-5910570 (pdf) Spiking neural network decoder for brain-machine interfaces

  • Golden standard: kalman filter.
  • Spiking neural network got within 1% of this standard.
  • THe 'neuromorphic' approach.
  • Used Nengo, freely available neural simulator.


Dethier, J. and Gilja, V. and Nuyujukian, P. and Elassaad, S.A. and Shenoy, K.V. and Boahen, K. Neural Engineering (NER), 2011 5th International IEEE/EMBS Conference on 396 -399 (2011)

hide / / print
ref: Kim-2006.06 tags: Hyun Kim Carmena Nicolelis continuous shared control gripper BMI date: 01-06-2012 00:20 gmt revision:2 [1] [0] [head]

IEEE-1634510 (pdf) Continuous shared control for stabilizing reaching and grasping with brain-machine interfaces.

  • The pneumatic gripper for picking up objects.
  • 70% brain control, 30% sensor control optimal.
  • Talk about 20Hz nyquist frequency for fast human motor movements, versus the need to smooth and remove noise.
  • Method: proximity sensors
    • collision avoidance 'pain withdrawal'
    • 'infant palmar grasp reflex'
    • Potential field associated with these sensors to implement continuous shared control.
  • Not! online -- used Aurora's data.


Kim, H.K. and Biggs, J. and Schloerb, W. and Carmena, M. and Lebedev, M.A. and Nicolelis, M.A.L. and Srinivasan, M.A. Continuous shared control for stabilizing reaching and grasping with brain-machine interfaces Biomedical Engineering, IEEE Transactions on 53 6 1164 -1173 (2006)

hide / / print
ref: tlh24-2011 tags: motor learning models BMI date: 01-06-2012 00:19 gmt revision:1 [0] [head]

Experiment: you have a key. You want that key to learn to control a BMI, but you do not want the BMI to learn how the key does things, as

  1. That is not applicable for when you don't have training data - amputees, parapalegics.
  2. That does not tell much about motor learning, which is what we are interested in.

Given this, I propose a very simple groupweight: one axis is controlled by the summed action of a certain population of neurons, the other by a second, disjoint, population; a third population serves as control. The task of the key is to figure out what does what: how does the firing of a given unit translate to movement (forward model). Then the task during actual behavior is to invert this: given movement end, what sequence of firings should be generated? I assume, for now, that the brain has inbuilt mechanisms for inverting models (not that it isn't incredibly interesting -- and I'll venture a guess that it's related to replay, perhaps backwards replay of events). This leaves us with the task of inferring the tool-model from behavior, a task that can be done now with our modern (though here-mentioned quite simple) machine learning algorithms. Specifically, it can be done through supervised learning: we know the input (neural firing rates) and the output (cursor motion), and need to learn the transform between them. I can think of many ways of doing this on a computer:

  1. Linear regression -- This is obvious given the problem statement and knowledge that the model is inherently linear and separable (no multiplication factors between the input vectors). n matlab, you'd just do mldivide (backslash opeartor) -- but but! this requires storing all behavior to date. Does the brain do this? I doubt it, but this model, for a linear BMI, is optimal. (You could extend it to be Bayesian if you want confidence intervals -- but this won't make it faster).
  2. Gradient descent -- During online performance, you (or the brain) adjusts the estimates of the weights per neuron to minimize error between observed behavior and estimated behavior (the estimated behavior would constitute a forward model..) This is just LMS; it works, but has a exponential convergence and may get stuck in local minima. This model will make predictions on which neurons change relevance in the behavior (more needed for acquiring reward) based on continuous-time updates.
  3. Batched Gradient descent -- Hypothetically, one could bolster the learning rate by running batches of data multiple times through a gradient descent algorithm. The brain very well could offline (sleep), and we can observe this. Such a mechanism would improve performance after sleep, which has been observed behaviorally in people (and primates?).
  4. Gated Gradient Descent -- This is halfway between reinforcement learning and gradient descent. Basically, the brain only updates weights when something of motivational / sensory salience occurs, e.g. juice reward. It differs from raw reinforcement learning in that there is still multiplication between sensory and motor data + subsequent derivative.
  5. Reinforcement learning -- Neurons are 'rewarded' at the instant juice is delivered; they adjust their behavior based on behavioral context (a target), which presumably (given how long we train our keys), is present in the brain at the same time the cursor enters the target. Sensory data and model-building are largely absent.

{i need to think more about model-building, model inversion, and songbird learning?}

hide / / print
ref: Kim-2007.08 tags: Hyun Kim muscle activation method BMI model prediction kinarm impedance control date: 01-06-2012 00:19 gmt revision:1 [0] [head]

PMID-17694874[0] The muscle activation method: an approach to impedance control of brain-machine interfaces through a musculoskeletal model of the arm.

  • First BMI that successfully predicted interactions between the arm and a force field.
  • Previous BMIs are used to decode position, velocity, and acceleration, as each of these has been shown to be encoded in the motor cortex
  • Hyun talks about stiff tasks, like writing on paper vs . pliant tasks, like handling an egg; both require a mixture of force and position control.
  • Georgopoulous = velocity; Evarts = Force; Kalaska movement and force in an isometric task; [17-19] = joint dependence;
  • Todorov "On the role of primary motor cortex in arm movement control" [20] = muscle activation, which reproduces Georgouplous and Schwartz ("Direct cortical representation of drawing".
  • Kakei [19] "Muscle movement representations in the primary motor cortex" and Li [23] [1] show neurons correlate with both muscle activations and direction.
  • Argues that MAM is the best way to extract impedance information -- direct readout of impedance requires a supervised BMI to be trained on data where impedance is explicitly measured.
  • linear filter does not generalize to different force fields.
  • algorithm activity highly correlated with recorded EMG.
  • another interesting ref: [26] "Are complex control signals required for human arm movements?"


[0] Kim HK, Carmena JM, Biggs SJ, Hanson TL, Nicolelis MA, Srinivasan MA, The muscle activation method: an approach to impedance control of brain-machine interfaces through a musculoskeletal model of the arm.IEEE Trans Biomed Eng 54:8, 1520-9 (2007 Aug)
[1] Li CS, Padoa-Schioppa C, Bizzi E, Neuronal correlates of motor performance and motor learning in the primary motor cortex of monkeys adapting to an external force field.Neuron 30:2, 593-607 (2001 May)

hide / / print
ref: Velliste-2008.06 tags: Schwartz 2008 Velliste BMI feeding population vector date: 01-06-2012 00:19 gmt revision:1 [0] [head]

PMID-18509337[0] Cortical control of a prosthetic arm for self-feeding

  • Idea: move BMI into robotic control.
  • population vector control, which has been shown to be inferior to the Wiener filter.
  • 112 units for control in one monkey. 2 monkeys used.
  • 4D control -- x, y, z, gripper.
  • 1064 trials over 13 days, average success rate of 78%
  • Gripper opened as the arm returned to mouth. Works b/c marshmallows are sticky.


[0] Velliste M, Perel S, Spalding MC, Whitford AS, Schwartz AB, Cortical control of a prosthetic arm for self-feeding.Nature 453:7198, 1098-101 (2008 Jun 19)

hide / / print
ref: Peikon-2009.06 tags: Peikon Fitzsimmons Nicolelis video tracking walking BMI Idoya date: 01-06-2012 00:19 gmt revision:2 [1] [0] [head]

PMID-19464514[0] Three-dimensional, automated, real-time video system for tracking limb motion in brain-machine interface studies.

  • yepp.


[0] Peikon ID, Fitzsimmons NA, Lebedev MA, Nicolelis MA, Three-dimensional, automated, real-time video system for tracking limb motion in brain-machine interface studies.J Neurosci Methods 180:2, 224-33 (2009 Jun 15)

hide / / print
ref: Fagg-2007.1 tags: BMI kinarm Hatsopoulos Moxon Miller FES date: 01-06-2012 00:17 gmt revision:5 [4] [3] [2] [1] [0] [head]

PMID-17978021[0] Biomimetic Brain Machine Interfaces for the Control of Movement.

  • images/482_1.pdf
  • describe structured models that include arm information & 'plant' dynamics.
    • current methods ignore the dynamics of the musculoskeletal system. Want to mimic natural arm movement.
    • To this end used a kinarm with a paralyzed monkey.
  • obtained real-time prediction of joint force, torque, and EMG
    • Concerning quality of prediction: they use fraction of movement variance that can be accounted for (FVAF) which, though google does not seem to know much about it, is probably the same as R^2. but it does not look that great:
      • 0.61 - 0.65 for torque prediction
      • 0.70 - 0.75 for EMG prediction once again, the limitation is the recording technology.
  • tested coupling predictions to the freehand FES system - see this crazy news brief
  • want to incorporate somatosensory feedback into the BMI.
  • they reference a paper from 2008 - huh? The document claims to be written/published in 2007.


[0] Fagg AH, Hatsopoulos NG, de Lafuente V, Moxon KA, Nemati S, Rebesco JM, Romo R, Solla SA, Reimer J, Tkach D, Pohlmeyer EA, Miller LE, Biomimetic brain machine interfaces for the control of movement.J Neurosci 27:44, 11842-6 (2007 Oct 31)

hide / / print
ref: Brown-2007 tags: Kalman filter BMI Black spike_sorting Donoghue date: 01-06-2012 00:07 gmt revision:1 [0] [head]

From Uncertain Spikes to Prosthetic Control a powerpoint presentation w/ good overview of all that the Brown group has done

hide / / print
ref: notes-2007 tags: clementine BMI robot kinarm timarm 032807 date: 01-06-2012 00:07 gmt revision:14 [13] [12] [11] [10] [9] [8] [head]

  1. http://m8ta.com/tim/clementine.MOV -- opens with totem, MJPG compressor.
  2. http://m8ta.com/tim/timarm_servocontroller.JPG
  3. http://m8ta.com/tim/images/spikeInformation_shuffled.jpg
    1. shuffled information distribution -- high significance level ;)
  4. kinarm.
    1. http://www.hardcarve.com/tim/kinarm.JPG
    2. http://www.hardcarve.com/tim/kinarm2.JPG
    3. http://www.hardcarve.com/tim/kinarm3.JPG
  5. robot svg or timarm png
    1. http://www.hardcarve.com/tim/timarm/timarm_side.jpg
    2. http://m8ta.com/tim/robotPulleyDetail.png
  6. bmi predictions clem 032807
      1. x & y predictions
      1. x & y predictions
      1. z velocity predictions - pretty darn good, snr 2
    1. Movie of the day: http://m8ta.com/tim/clem032807_3dBMI.MPG
      1. cells for that day - 40 in all

hide / / print
ref: BASMAJIAN-1963.08 tags: original BMI M1 human EMG tuning operant control Basmajian date: 01-05-2012 00:49 gmt revision:6 [5] [4] [3] [2] [1] [0] [head]

PMID-13969854[0] Control and Training of Individual Motor Units

  • humans have the ability to control the firing rate of peripheral motor units with a high resolution.
  • "The quality of control over anterior horn cells may determine the rates of learning" yup!
  • "Some learn such esquisite control that they soon can produce rhythms of contraction in one unit, imitating drum rolls etc"
  • the youngest persons were among both the best and worst learners.
  • after about 30 minutes the subject was required to learn how to repress the first unit and to recruit another one.
    • motor unit = anterior horn cell, its axon, and all the muscle fibers on which the terminal branches of the axon end. max rate ~= 50hz.
    • motor units can be discriminated, much like cortical neurons, by their shape.
    • some patients could recruit 3-5 units altogether - from one bipolar electrode!
      • in playback mode (task: trigger the queried unit), several subjects had particular difficulty in recruiting the asked-for units. "They groped around in their conscious efforts to find them sometimes, it seemed, only succeded by accident"
    • some patients could recruit motor units in the absence of feedback, but they were unable to explain how they do it.
  • 0.025 (25um) nylon-insulated Karma alloy EMG recording wire.
  • feedback: auditory & visual (oscilloscope).
  • motor units have a maximum rate, above which overflow takes place and other units are recruited (in accord with the size principle).
  • "The controls (are) learned so quickly, are so esquisite, are so well retained after the feedbacks are eliminated that one must not dismiss them as tricks"


hide / / print
ref: Wyler-1974.02 tags: Wyler Fetz BMI operant conditioning date: 01-05-2012 00:46 gmt revision:3 [2] [1] [0] [head]

PMID-4207598[0] Behavioral control of firing patterns of normal and abnormal neurons in chronic epileptic cortex.

  • Idea: epilepsy treated through biofeedback.
  • Induced epilepsy in monkeys via alumina.
  • Conditioned 198 cells in epileptiform focus; 107 had normal firing patterns.
  • 91 cells had abnormal patterns:
    • Structured bursts with high, invariant burst indices, and could not be conditioned.
    • Cells did not change burstyness based on behavioral state.
    • Lower and more variable burst indices and were as easily conditioned as normal cells.
      • These cells bursted more when the monkey was not paying attention.
  • Operant control: ref 8, 9.
  • Ach, fascinating:
  • Normal precentral cells rarely exhibited interspike intervals less than 10 msec, except during vigorous movements or sleep.
  • Neurons were deemed 'bursty' if they exhibited spontaneous high-frequency firing with interspike intervals less that 5msec.
  • Monkeys obtained proficiency with high-frequency conditioning more quickly and effectively than with low-freq, even with 40% on high and 60% on low.
  • All conditioned cells corresponded to some movement of the contralateral arm (again).
  • Operant conditioning is interesting in this case, as it indicates if cells are still 'functional' in the ensemble.
  • See also: PMID-809116[1]


[0] Wyler AR, Fetz EE, Behavioral control of firing patterns of normal and abnormal neurons in chronic epileptic cortex.Exp Neurol 42:2, 448-64 (1974 Feb)
[1] Wyler AR, Fetz EE, Ward AA Jr, Firing patterns of epileptic and normal neurons in the chronic alumina focus in undrugged monkeys during different behavioral states.Brain Res 98:1, 1-20 (1975 Nov 7)

hide / / print
ref: Schmidt-1980.01 tags: BMI 1980 SUA M1 prosthetics Schmidt MEA date: 01-04-2012 22:59 gmt revision:14 [13] [12] [11] [10] [9] [8] [head]

PMID-6794389[0] Single neuron recording from motor cortex as a possible source of signals for control of external devices

  • also [1]
  • I guess this was the first published article claiming that motorneurons could be used to drive a prosthesis, and first clear attempt at long-term array recording (?)
  • recorded via arrays for up to 37 months!
    • only 2 of the 11 eelctrodes were recording at the time of sacrifice.
  • trained the monkey to perform an 8 target tracking task
    • with cortical neurons: 2.45 bits/second
    • with wrist flexion/extension: 4.48 bits/second
  • electrodes: {946} A new chronic recording intracortical microelectrode (1976!)
    • 25um iridium wires electropolished to a 1um tip; 1.5mm long.
    • electrodes float on the cortex; signals transmitted through 25um gold wire, which is in turn connected to a head-mounted connector.
    • iridium and gold are insulated with vapor-deposited parylene-C
    • electrode tips are exposed with a HV arc. (does this dull them? from the electromicrograph, it seems that it just makes them rougher.)
    • arrays of 12.
    • 1M impedance (average)
  • interesting: neural activity was recorded from at least 8 different neurons with this electrode during the course of the implant, indicating that it was migrating through cortical tissue.
    • the average recording time from the same electrode was 8 days; max 23 days.
  • second implant was more successful: maximium time recording from the same neuron was 108 days.
  • failure is associated with cracks in the parylene insulation (which apparently occurred on the grain boundaries of the iridium). "still only marginally reliable" (and still.. and still..)
  • they have operantly trained cortical units in another, earlier study.
  • have, effectively, 8 levels of activity, with feedback monkey has to match the proscribed firing rate.
  • > 50% rewarded trials = success for them; 26/28 of the neurons tested were eventually conditioned successfully.
  • looks like the monkey can track the target firing rate rather accurately. "the output of cortical cells can provide information output rates moderately less precise than the intact motor system. "
  • Monkey can also activated sequences of neurons: A, then AB, then B.
  • people have also tried conditioning individual EMG units; it is sometimes possible to control 2 different motor units in the same muscle independently, but in general only a single channel of information can be obtained from one muscle, and gross EMGs are fine for this.
    • Thus surface EMG is preferred.
    • you can get ~ 2.73 bits/sec with gross EMG on a human; 2.99 bits/sec (max) with a monkey.
  • they remind us, of course, that an enormous amount of work remains to be done.


hide / / print
ref: Olson-2005 tags: Arizona rats BMI motor control training SVM single-unit left right closed-loop learning Olson Arizona date: 01-03-2012 23:06 gmt revision:1 [0] [head]

bibtex:Olson-2005 Evidence of a mechanism of neural adaptation in the closed loop control of directions

  • from abstract:
    • Trained rats to press left/right paddles to center a LED. e.g. paddles were arrow keys, LED was the cursor, which had to be centered. Smart rats.
      • Experiment & data from Olson 2005
    • Then trained a SVM to discriminate left/right from 2-10 motor units.
    • Once closed-loop BMI was established, monitored changes in the firing properties of the recorded neurons, specifically wrt the continually(?) re-adapted decoding SVM.
    • "but expect that the patients who use the devices will adapt to the devices using single neuron modulation changes. " --v. interesting!
  • First page of article has an excellent review back to Fetz and Schmidt. e.g. {303}
  • Excellent review of history altogether.
    • Notable is their interpretation of Sanchez 2004 {259}, who showed that most of the significant modulations are from a small group of neurons, not the large (up to 320 electrodes) populations that were actually recorded. Carmena 2003 showed that the population as a whole tended to group tuning, although this was imperfectly controlled.
  • Also reviewed: Zacksenhouse 2007 {901}
  • SVM is particularly interesting as a decoding algorithm as it weights the input vectors in projecting onto a decision boundary; these weights are experimentally informative.
  • Figure 7: The brain seems to modulate individual firing rate changes to move away from the decision boundary, or at least to minimize overlap.
  • For non-overt movements, the distance from decision function was greater than for overt movements.
  • Rho ( ρ\rho ) is the Mann-Whitney test statistic, which non-parametrically estimates the difference between two distributions.
  • δf(X t)\delta f(X_t) is the gradient wrt the p input dimensions o9f the NAV, as defined with their gaussian kernel SVM.
  • They show (i guess) that changes in ρ\rho are correlated with the gradient -- e.g. the brain focuses on neurons that increase fidelity of control?
    • But how does the brain figure this out??
  • Not sure if i fully understand their argument / support.
  • Conclusion comes early in the paper
    • figure 5 weakly supports the single-neuron modulation result.

hide / / print
ref: Hatsopoulos-2005.01 tags: BMI Hatsopoulos Donoghue cortex date: 01-03-2012 22:49 gmt revision:4 [3] [2] [1] [0] [head]

PMID-17282055[0][] Cortically controlled brain-machine interface

  • conference proceedings. describe the 6month teraplegic trial.
  • (above, monkey)
    • lets them record from 40% of electrodes.
    • 100-200uv units, 20uv noise.
    • one year to three years post implantation.
  • advocate hybrid multimodal control.
    • M1 = continuous control
    • PMd = discrete control
      • used a probabilistic model for this (poisson firing rate, individual neurons are independent)


[0] Hatsopoulos N, Mukand J, Polykoff G, Friehs G, Donoghue J, Cortically controlled brain-machine interface.Conf Proc IEEE Eng Med Biol Soc 7:1, 7660-7663 (2005)

hide / / print
ref: neuro-2005 tags: NRSA background BMI tool use date: 01-03-2012 15:21 gmt revision:2 [1] [0] [head]

  • tool use:
    • [0]
    • [1] varying neural responses following tool acquisition
  • BMI
    • [2] simultaneous prediction of 4 variables
  • spike sorting
    • [3] donoghue
    • [4] LFP
    • [5] MUA
    • [6,7] - 1980!!
    • [8] STN bmi (nahh)
    • [9] Shenoy, eye movement better, 6.5 bits/sec
    • [10] PF
    • [11] in rats, in the cinglate, still they didn't get the point.
    • [12] Fetz stimulation


hide / / print
ref: ODoherty-2009.01 tags: Odoherty Nicolelis ICMS stimulation BMI BMBI date: 01-03-2012 06:55 gmt revision:1 [0] [head]

PMID-19750199[0] A brain-machine interface instructed by direct intracortical microstimulation.

  • Both and and brain control, cud by ICMS to S1, Mango and Nectarine.
    • PP ineffective. Despite Doty [1].
  • pretty careful site mapping (fig 1).
  • SUA classified by less that 1 per 1000 ISI < 1.6ms.
  • pursuit & center out tasks.
  • Correlation coeficient (R^2) not so high across all sessions - 0.5 (?).
  • ICMS learning, once the monkey began to get it, was rapid.


[0] O'Doherty JE, Lebedev MA, Hanson TL, Fitzsimmons NA, Nicolelis MA, A brain-machine interface instructed by direct intracortical microstimulation.Front Integr Neurosci 3no Issue 20 (2009)
[1] Doty RW, Electrical stimulation of the brain in behavioral context.Annu Rev Psychol 20no Issue 289-320 (1969)

hide / / print
ref: Lebedev-2006.09 tags: misha lebedev nicolelis BMI past present future electrodes review MEA date: 01-03-2012 03:26 gmt revision:2 [1] [0] [head]

PMID-16859758 Brain-machine interfaces: past, present and future.

  • Describes the state of the art & what needs to be done. Namely, better recording quality.
  • state that current EEG BMIs are limited to 5-25 bits/min (typo!!) [2,11]
    • [2] Wolpaw "Brain-computer interfaces for communication and control" 2002.
    • [11] Birbaumer "Brain-computer-interface research: coming of age" 2006.
  • set of references on biofeedback control of EEG in research animals.
  • EEG BCIs are either biofeedback based or classifier (P300, synchronous decoder ) based.
  • First invasive BMIs Fetz [40-45], Schmidt [46]; progress has been limited by technology. Must have been quite challenging for him to do the research!
    • [40] Fetz "Operant Conditioning of cortical unit activity" 1969
    • [41] Fetz "Are movement parameters recognizably coded in activity of single neurons?" 1992
    • [42] Fetz and Baker "Operantly conditioned patterns on precentral unit activity and correlated responses in adjacent cells and contralateral muscles." 1973
    • [43] Fetz and Finocchio "Operant Conditioning of specific patterns of neural and muscular activity" 1971
    • [44] Fetz and Finocchio "Operant conditioning of isolated activity in specific muscles and precentral cells" 1972
    • [45] Fetz and Finocchio 1975 "Correlations between activity of motor cortex cells and arm muscles during operantly conditioned response patterns." 1975
    • [46] Schmidt "single neuron recording from motor cortex as a possible source of signals for control of external devices." 1980
  • microelectrode arrays solved one of the problems.
  • talk about how more neurons are needed.
  • Principles of BMIs: Evarts [66-68], neuronal modulations are highly variable [69-72].
    • [66] Evarts, E.V. (1966) Pyramidal tract activity associated with a conditioned hand movement in the monkey.
    • [67] Evarts, E.V. (1968) Relation of pyramidal tract activity to force exerted during voluntary movement.
    • [68] Evarts, E.V. (1968) A technique for recording activity of subcortical neurons in moving animals.
    • "THus, as much as neighboring neurons might display highly disinct firing modulation patterns during the execution of a particular movement, single-neuron firing can vary substantially from one trial to the next, despite the fact that the overt movements remain virtually identical. :
    • "averaging across large populations of neurons significantly reduces the variability of signals derived from single neurons [54, 69].
    • Should i mention this in thesis?
  • Better way to assimilate the BMI into the body is to have proprioceptive feedback.
  • suggest the same standard things to be improved, excluding electronics. :
    • electrodes / recording
    • decoding
    • incorporating plasticity
    • better prosthetics.
  • "multi-unit signals can also be efficiently used in BMI control [57] {318}.
  • Some groups have strongly claimed that recordings from a small number of neurons can be sufficient for good performance in a BMI. [55,56,63]
    • This is not Miguel's approach: more neurons confers accuracy [54{317},57,70] and reliability [69].
      • [70] Wessberg, J. and Nicolelis, M.A. (2004) Optimizing a linear algorithm for real-time robotic control using chronic cortical ensemble recordings in monkeys. J. Cogn. Neurosci. 16, 1022–1035
  • Still need new microelectrodes; electrodes become encapsulated by fibrous tissue and cells die in the vicinity of electrodes [77] {781}.
    • suggest anti-inflammatory coating, though the jury is out.
  • Initial wireless telemetry systems: [93-99]. [93]{315}
    • [94] Knutti, J.W. et al. (1979) An integrated circuit approach to totally implantable telemetry systems. Biotelem. Patient Monit. 6, 95–106
    • [97] Chien, C.N. and Jaw, F.S. (2005) Miniature telemetry system for the recording of action and field potentials. J. Neurosci. Methods 147
    • [98] {930}
    • [99] Morizio Morizio, J. et al. (2005) Fifteen-channel wireless headstage system for single-unit rat recordings.
    • [100] (of broader interest) Moxon, K.A. et al. (2004) Ceramic-based multisite electrode arrays for chronic single-neuron recording. IEEE Trans. Biomed. Eng. 51, 647–656
  • nanotechnology probes that access the brain through the vasular system [101].
  • although a good number of linear and nonlinear algorithms have been proposed and tested [1,54,56,57,70,110-116], Wiener filters have proved sufficient [54,55,57,58,65,117].
  • almost 140 years ago Head and Holmes suggested that the body schema -- that is, the internal brain representation of one's body -- could extend itself to include a wielded tool.

hide / / print
ref: Rolston-2009.01 tags: ICMS artifacts stimulation Rolston Potter recording BMI date: 01-03-2012 02:38 gmt revision:3 [2] [1] [0] [head]

PMID-19668698[0] A low-cost multielectrode system for data acquisition enabling real-time closed-loop processing with rapid recovery from stimulation artifacts

  • Well written, well tested, but fundamentally simple system - only two poles active high-pass, one pole low-pass.
  • With TBSI headstages the stimulation artifact is brief - figure 8 shows < 4ms.
  • Includes NeuroWriter software, generously open-sourced (but alas windows only - C#).


[0] Rolston JD, Gross RE, Potter SM, A low-cost multielectrode system for data acquisition enabling real-time closed-loop processing with rapid recovery from stimulation artifacts.Front Neuroengineering 2no Issue 12 (2009)

hide / / print
ref: Darmanjian-2005.03 tags: recording wifi 802.11 DSP BMI Principe date: 01-03-2012 02:13 gmt revision:2 [1] [0] [head]

IEEE-1419566 (pdf) A Portable Wireless DSP System for a Brain Machine Interface

  • 1400Mw (yuck!!), large design, PCMCIA 802.11 card @ 1.8 Mbps, external SRAM for models
  • implemented LMS and as expected it's faster on the Texas Instruments C33 floating-point DSP.


Darmanjian, S. and Morrison, S. and Dang, B. and Gugel, K. and Principe, J. Neural Engineering, 2005. Conference Proceedings. 2nd International IEEE EMBS Conference on 112 -115 (2005)

hide / / print
ref: Sanchez-2005.06 tags: BMI Sanchez Nicolelis Wessberg recurrent neural network date: 01-01-2012 18:28 gmt revision:2 [1] [0] [head]

IEEE-1439548 (pdf) Interpreting spatial and temporal neural activity through a recurrent neural network brain-machine interface

  • Putting it here for the record.
  • Note they did a sensitivity analysis (via chain rule) of the recurrent neural network used for BMI predictions.
  • Used data (X,Y,Z) from 2 monkeys feeding.
  • Figure 6 is strange, data could be represented better.
  • Also see: IEEE-1300786 (pdf) Ascertaining the importance of neurons to develop better brain-machine interfaces Also by Justin Sanchez.


Sanchez, J.C. and Erdogmus, D. and Nicolelis, M.A.L. and Wessberg, J. and Principe, J.C. Interpreting spatial and temporal neural activity through a recurrent neural network brain-machine interface Neural Systems and Rehabilitation Engineering, IEEE Transactions on 13 2 213 -219 (2005)

hide / / print
ref: Carmena-2005.11 tags: carmena BMI nicolelis single-unit variability 2005 date: 01-01-2012 17:31 gmt revision:2 [1] [0] [head]

PMID-16291944[0] Stable ensemble performance with single-neuron variability during reaching movements in primates.

  • correlation between the firing of single neurons and movement parameters was nonstationary over 30-60 minute recording sessions.
  • yet! you could get stable prediction of arm movements, suggesting that movement parameters are redundantly encoded.
  • this, in turn, implies that you do not need a stable recorded population for good predictions.
  • suggest that the variance itself could be a means of neuronal 'computation' or exploration based on perturbations.
    • later Carmena papers do not mention this.


hide / / print
ref: Obeid-2004.02 tags: Wolf BMI recording electronics telemetry Obeid date: 12-31-2011 18:27 gmt revision:4 [3] [2] [1] [0] [head]

PMID-14757341[1] A low power multichannel analog front end for portable neural signal recordings.

  • have an interesting section on CMRR, quote: Although we use a precision differential amplifier with a CMRR of 110 dB, we were unable, in practice, to measure CMRRs greater than not, vert, similar42 dB. This can be accounted for by the device tolerances in the preamplifier stage; using ±0.1% resistors and ±5% capacitors in the preamplifier, the expected worst case CMRR at 1 kHz is 39.2 dB


hide / / print
ref: OLDS-1954.12 tags: Olds Milner operant conditioning electrical reinforcement wireheading BMI date: 12-29-2011 05:09 gmt revision:5 [4] [3] [2] [1] [0] [head]

PMID-13233369[0] Positive reinforcement produced by electrical stimulation of septal area and other regions of rat brain.

  • The original electrical reinforcement experiment!
  • tested out various areas for reinforcement; septal forebrain area was the best.
  • later work: 1956 Olds, J. Runway and maze behavior controlled by basomedial forebrain stimulation in the rat. J. Comp. Physiol. Psychol. 49:507-12.


[0] OLDS J, MILNER P, Positive reinforcement produced by electrical stimulation of septal area and other regions of rat brain.J Comp Physiol Psychol 47:6, 419-27 (1954 Dec)

hide / / print
ref: Vaadia-2009.09 tags: BMI Vaadia 2009 date: 12-28-2011 20:39 gmt revision:2 [1] [0] [head]

PMID-20228862[0] Grand Challenges of Brain Computer Interfaces in the Years to Come

  • Problem 1: If you have no theory of mind, you just keep making a series of measurements.
  • EEG is like listening to lectures of millions of people ... simultaneously.
  • Single unit recordings is like probing a microchip's individual wires and trying to figure out what it's doing (this is not a good analogy, though -- brains are far more robust to part failure than a computer).
  • References Todorov 2004 PMID-15332089[1] Wolpert and Ghahramani 2000: Sensorimotor control.
  • Problem 2: noisy measurements.
    • Might be the brains problem, not us: neuronal interactions modify rapidly during sensorimotor learning. Jarosiewicz et al 2008. PMID-19047633[2]
    • Claim to have a system that learns control within 1-2 minutes or '10 seconds', coadaptively. Shpigelman 2009. bibtex:Shpigelman-2009
  • In drug resistant focal epilepsy, not only were substantial reductions in seizures reported, but also large gains in IQ and cognitive functioning were demonstrated (Kotchoubey 2001, Strehl et al 2005) after training of slow cortical potential control. Better ref: PMID-10457815[3]
  • Buch et al 2008 Demonstrated MEG BMI control. PMID-18258825[4]
  • Even with sophisticated classification solutions, EEG cannot provide much better than 2D control (Birbaumer 1990). PMID-2404287[5]
  • Ref Moritz [6]
  • Supposes that nanotechnology may ultimately find a solution -- inert nanoprobes that measure activity and transmit a compressed version.


[0] Vaadia E, Birbaumer N, Grand challenges of brain computer interfaces in the years to come.Front Neurosci 3:2, 151-4 (2009 Sep 15)
[1] Todorov E, Optimality principles in sensorimotor control.Nat Neurosci 7:9, 907-15 (2004 Sep)
[2] Jarosiewicz B, Chase SM, Fraser GW, Velliste M, Kass RE, Schwartz AB, Functional network reorganization during learning in a brain-computer interface paradigm.Proc Natl Acad Sci U S A 105:49, 19486-91 (2008 Dec 9)
[3] Thompson L, Thompson M, Neurofeedback combined with training in metacognitive strategies: effectiveness in students with ADD.Appl Psychophysiol Biofeedback 23:4, 243-63 (1998 Dec)
[4] Buch E, Weber C, Cohen LG, Braun C, Dimyan MA, Ard T, Mellinger J, Caria A, Soekadar S, Fourkas A, Birbaumer N, Think to move: a neuromagnetic brain-computer interface (BCI) system for chronic stroke.Stroke 39:3, 910-7 (2008 Mar)
[5] Birbaumer N, Elbert T, Canavan AG, Rockstroh B, Slow potentials of the cerebral cortex and behavior.Physiol Rev 70:1, 1-41 (1990 Jan)
[6] Moritz CT, Perlmutter SI, Fetz EE, Direct control of paralysed muscles by cortical neurons.Nature 456:7222, 639-42 (2008 Dec 4)

hide / / print
ref: Dagnelie-2008.01 tags: visual BMI prosthesis review Dagnelie date: 12-17-2011 02:25 gmt revision:0 [head]

PMID-18429703 Psychophysical evaluation for visual prosthesis.

  • Visual prostheses are clinical and preclinical trials!
  • cochlear implants function with 16-20 electrodes; retina is 120e6 photoreceptors and 1.2 optic nerve fibers.
  • Argus 2 retinal implant has 60 electrodes. visual information impoverished.
  • In the heyday of prewar German scientific discovery, Foerster (3) established that electrical stimulation of the visual cortex in an awake patient during a neurosurgical intervention produced the percept of dots of light, called phosphenes, and that the location of a phosphene changed with that of the electrical stimulus.
  • people originally thought that loss of the photoreceptors would lead to degradation of the RGCs; this appears not to be true.
  • There is broad consensus that functional vision restoration is predicated on prior visual experience; this is different than cochlera prostheses, which work on congenitally deaf people.
    • Visual development depends on nearly a decade of high-resolution perception, and cannot be emulated later in life through a low-bw prosthesis.
  • There are at the present time at least 20 distinct research groups in at least 8 countries actively engaged in visual prosthesis development.
  • discuss a lot of pre-clinical testing & all the nitty-grity details, e.g. how to make a low res prosthesis work for reading.

hide / / print
ref: Guenther-2009.12 tags: Guenther Kennedy 2009 neurotrophic electrode speech synthesize formant BMI date: 12-17-2011 02:12 gmt revision:2 [1] [0] [head]

PMID-20011034[0] A Wireless Brain-Machine Interface for Real-Time Speech Synthesis

  • Neurites grow into the glass electrode over the course of 3-4 months; the signals and neurons are henceforth stable, at least for the period prior publication (>4 years).
  • Used an FM modulator to send out the broadband neural signal; powered the implanted electronics inductively.
  • Sorted 56 spike clusters (!!)
    • quote: "We chose to err on the side of overestimating the number of clusters in our BMI since our Kalman filter decoding technique is somewhat robust to noisy inputs, whereas a stricter criterion for cluster definition might leave out information-carrying spike clusters."
    • 27 units on one wire and 29 on the other.
  • Quote: "neurons in the implanted region of left ventral premotor cortex represent intended speech sounds in terms of formant frequency trajectories, and projections from these neurons to primary motor cortex transform the intended formant trajectories into motor commands to the speech articulators."
    • Thus speech can be represented as a trajectory through formant space.
    • plus there are many simple low-load formant-based sw synthesizers
  • Used supervised methods (ridge regression), where the user was asked to imagine making vowel sounds mimicking what he heard.
    • only used the first 2 vowel formants; hence 2D task.
    • Supervised from 8 ~1-minute recording sessions.
  • 25 real-time feedback sessions over 5 months -- not much training time, why?
  • Video looks alright.


[0] Guenther FH, Brumberg JS, Wright EJ, Nieto-Castanon A, Tourville JA, Panko M, Law R, Siebert SA, Bartels JL, Andreasen DS, Ehirim P, Mao H, Kennedy PR, A wireless brain-machine interface for real-time speech synthesis.PLoS One 4:12, e8218 (2009 Dec 9)

hide / / print
ref: Kennedy-1992.08 tags: BMI Kennedy cone electrode electrophysiology recording neurotrophic date: 12-17-2011 01:00 gmt revision:1 [0] [head]

PMID-1407726[] The cone electrode: ultrastructural studies following long-term recording in rat and monkey cortex

  • they placed sciatic nerve inside the glass cone electrode to encourage regrowth.
    • alternatively, they filled the cone electrode with 'matrigel' whatever that 'neurotrophic substance' is.
  • good recordings at 6 months post impantation.
  • virtually no neurons were found in the tissue in any cone
    • however, they saw plenty of mylenated axons. (the mylenation assuredly is good for the quality of recordings hah)
    • in no case was the tissue absent from the glass.


hide / / print
ref: Jackson-2006.11 tags: Fetz Andrew Jackson BMI motor learning microstimulation date: 12-16-2011 04:20 gmt revision:6 [5] [4] [3] [2] [1] [0] [head]

PMID-17057705 Long-term motor cortex plasticity induced by an electronic neural implant.

  • used an implanted neurochip.
  • record from site A in motor cortex (encodes movement A)
  • stimulate site B of motor cortex (encodes movement B)
  • after a few days of learning, stimulate A and generate mixure of AB then B-type movements.
  • changes only occurred when stimuli were delivered within 50ms of recorded spikes.
  • quantified with measurement of (to) radial/ulnar deviation and flexion/extension of the wrist.
  • stimulation in target (site B) was completely sub-threshold (40ua)
  • distance between recording and stimulation site did not matter.
  • they claim this is from Hebb's rule: if one neuron fires just before another (e.g. it contributes to the second's firing), then the connection between the two is strengthened. However, i originally thought this was because site A was controlling the betz cells in B, therefore for consistency A's map was modified to agree with its /function/.
  • repetitive high-frequency stimulation has been shown to expand movement representations in the motor cortex of rats (hmm.. interesting)
  • motor cortex is highly active in REM


hide / / print
ref: Patil-2004.07 tags: BMI DBS Parag Patil Carmena Turner date: 12-14-2011 23:45 gmt revision:4 [3] [2] [1] [0] [head]

PMID-15214971[0] Ensemble recordings of human subcortical neurons as a source of motor control signals for a brain-machine interface

  • they do not show ISI or autocorrelations functions for any of the neurons. however, some of the continuous recordings look really *excellent*.
  • there are a lot of comments at the end of the paper, may which are quite intelligent & informative.
    • motor control becomes defocused in Parkinsons, e.g. too many motor units respond to a movement. This is reduced with the use of dopamine agonists like atrophine.
  • Parag Patil's Homepage


hide / / print
ref: Won-2004.02 tags: Debbie Won Wolf spike sorting mutual information tuning BMI date: 12-07-2011 02:58 gmt revision:3 [2] [1] [0] [head]

PMID-15022843[0] A simulation study of information transmission by multi-unit microelectrode recordings key idea:

  • when the units on a single channel are similarly tuned, you don't loose much information by grouping all spikes as coming from one source. And the opposite effect is true when you have very differently tuned neurons on the same channel - the information becomes more ambiguous.


hide / / print
ref: Wyler-1985.08 tags: Wyler synchrony operant conditioning BMI date: 12-06-2011 06:36 gmt revision:1 [0] [head]

PMID-4041789 Synchrony between cortical neurons during operant conditioning.

  • Fetz and Baker showed that individual neurons recorded from the same electrode can modulate their firing upon operant conditioning either together, opposite, or independently.
  • Wyler has duplicated this result, and undertakes this further analysis to show that these pairs of neurons recorded from the same electrode show high degrees (67%) of tight 1ms synchrony.
  • This despite the fact that in 80% of cases the firing rates did not covary.
  • This suggests that they must have a common synaptic pathway.
  • Reference (and support) Lemon and Porter's finding that adjacent neurons respond to widely separated peripheral fields.

hide / / print
ref: Helms-2003.01 tags: Schwartz BMI adaptive control Taylor Tillery 2003 date: 11-26-2011 00:58 gmt revision:1 [0] [head]

PMID-12929922 Training in cortical control of neuroprosthetic devices improves signal extraction from small neuronal ensembles.

  • Lays out the coadaprive algorithm.
  • with supervised / adaptive training, ML estimator is able to get 80% of the targets correct.
  • Reviews in the Neurosciences (conference) Workshop on Neural and Artificial Computation.

hide / / print
ref: Chapin-1999.07 tags: chapin Nicolelis BMI neural net original SUNY rat date: 09-02-2009 23:11 gmt revision:2 [1] [0] [head]

PMID-10404201 Real-time control of a robot arm using simultaneously recorded neurons in the motor cortex.

  • Abstract: To determine whether simultaneously recorded motor cortex neurons can be used for real-time device control, rats were trained to position a robot arm to obtain water by pressing a lever. Mathematical transformations, including neural networks, converted multineuron signals into 'neuronal population functions' that accurately predicted lever trajectory. Next, these functions were electronically converted into real-time signals for robot arm control. After switching to this 'neurorobotic' mode, 4 of 6 animals (those with > 25 task-related neurons) routinely used these brain-derived signals to position the robot arm and obtain water. With continued training in neurorobotic mode, the animals' lever movement diminished or stopped. These results suggest a possible means for movement restoration in paralysis patients.
The basic idea of the experiment. Rat controlled the water lever with a forelimb lever, then later learned to control the water lever directly. They used an artificial neural network to decode the intended movement.

hide / / print
ref: Radhakrishnan-2008.1 tags: EMG BMI Jackson motor control learning date: 10-03-2008 16:45 gmt revision:0 [head]

PMID-18667540[0] Learning a novel myoelectric-controlled interface task.

  • EMG-controlled 2D cursor control task with variable output mapping.
  • Subjects could learn non-intuitive output transforms to a high level of performance,
  • Subjects preferred, and learned better, if hand as opposed to arm muscles were used.


hide / / print
ref: bookmark-2006.07 tags: BMI BCI EEG bibliography Stephan Scott date: 09-07-2008 19:54 gmt revision:2 [1] [0] [head]



hide / / print
ref: Narayanan-2005.04 tags: Laubach M1 motor rats statistics BMI prediction methods date: 09-07-2008 19:51 gmt revision:4 [3] [2] [1] [0] [head]

PMID-15858046[] Redundancy and Synergy of Neuronal Ensembles in Motor Cortex

  • timing task.
  • rats.
  • 50um teflon microwires in motor cortex
  • ohno : neurons that were the best predictors of task performance were not necessarily the neurons that contributed the most predictive information to an ensemble of neurons.
  • most all contribute redundant predictive information to the ensemble.
    • this redundancy kept the predictions high, even if neurons were dropped.
  • small groups of neurons were more synergistic
  • large groups were more redundant.
  • used wavelet based discriminant pursuit.
    • validated with draws from a random data set.
  • used R and Weka
  • data looks hella noisy ?


hide / / print
ref: Serruya-2002.03 tags: BMI Donoghue 2002 Hatsopoulos Utah array Serruya date: 09-07-2008 19:08 gmt revision:1 [0] [head]

PMID-11894084[0] Instant neural control of a movement signal.

  • used only a few (7-30) motor cortex neurons
  • this let the monkey immediately manipulate a computer cursor, without extensive training (according to them).


hide / / print
ref: Fetz-2007.03 tags: hot fetz BMI biofeedback operant training learning date: 09-07-2008 18:56 gmt revision:7 [6] [5] [4] [3] [2] [1] [head]

PMID-17234689[0] Volitional control of neural activity: implications for brain-computer interfaces (part of a symposium)

  • Limits in the degree of accuracy of control in the latter studies can be attributed to several possible factors. Some of these factors, particularly limited practice time, can be addressed with long-term implanted BCIs. YES.
  • Accurate device control under diverse behavioral conditions depends significantly on the degree to which the neural activity can be volitionally modulated. YES again.
  • neurons (50%) in somatosensory (post central) cortex fire prior to volitional movement. interesting.
  • It should also be noted that the monkeys activated some motor cortex cells for operant reward without ever making any observed movements See: Fetz & Finocchio, 1975, PMID-810359.
    • Motor cortex neurons that were reliably associated with EMG activity in particular forelimb muscles could be readily dissociated from EMG when the rewarded pattern involved cell activity and muscle suppression.
    • This may be realated to switching between real and imagined movements.
  • Biofeedback worked well for activating low-threshold motor units in isolation, but not high threshold units; attempts to reverse recruitment order of motor units largely failed to demonstrate violations of the size principle.
  • This (the typical BMI decoding strategy) interposes an intermediate stage that may complicate the relationship between neural activity and the final output control of the device
    • again, in other words: "First, the complex transforms of neural activity to output parameters may complicate the degree to which neural control can be learned."
    • quote: This flexibility of internal representations (e.g. to imagine moving your arm, train the BMI on that, and rapidly directly control the arm rather than gonig through the intermediate/training step) underlies the ability to cognitively incorporate external prosthetic devices in to the body image, and explains the rapid conceptual adaptation to artificial environments, such as virtual reality or video games.
      • There is a high flexibility of input (sensory) and output (motor) for purposes of imagining / simulating movements.
  • adaptive learning algorithms may create a moving target for the robust learning algorithm; does it not make more sense to allow the cortex to work it's magic?
  • Degree of independent control of cells may be inherently contrained by ensemble interactions
    • To the extent that internal representations depend on relationships between the activities of neurons in an ensemble, processing of these representations involves corresponding constraints on the independence of those activities.
  • quote: "These factors suggest that the range and reliability of neural control in BMI might increase significantly when prolonged stable recordings are acheived and the subject can practice under consistent conditions over extended periods of time.
  • Fetz agrees that the limitation is the goddamn technology. need to fix this!
  • there is evidence of favortism in his citations (friends with Miguel??)

humm.. this paper came out a month ago, and despite the fact that he is much older and more experienced than i, we have arrived at the same conclusions by looking at the same set of data/papers. so: that's good, i guess.


hide / / print
ref: Ojakangas-2006.12 tags: BMI Donoghue prosthetic DBS prefrontal cortex planning date: 04-09-2007 22:32 gmt revision:3 [2] [1] [0] [head]

PMID-17143147[0] Decoding movement intent from human premotor cortex neurons for neural prosthetic applications

  • they suggest using additional frontal areas beyond M1 to provide signal sources for human neuromotor prosthesis.
    • did recording in prefrontal cortex during DBS surgeries.
    • these neurons were able to provide information about movement planning production, and decision-making.
  • unusual for BMI studies, their significance levels are near 0.02 - they show distros of % correct based on a ML decoding scheme.


hide / / print
ref: Humphrey-1970.11 tags: BMI original Humphrey date: 04-09-2007 19:47 gmt revision:1 [0] [head]

PMID-4991377[0] Predicting measures of motor performance from multiple cortical spike trains.

Recordings have been obtained simultaneously from several, individually selected neurons in the motor cortex of unanesthetized monkey as the animal performed simple arm movements. With the use of comparatively simple quantitative procedures, the activity of small sets of cells was found to be adequate for rather accurate real-time prediction of the time course of various response measurements. In addition, the results suggest that hypotheses concerning the response variables "controlled" by cortical motor systems may well depend upon whether or not the temporal relations between simultaneously active neurons are taken into account.

cited in miguel's book, "Methods for Neural ensemble recordings". However, I can't get the text online.


hide / / print
ref: Birbaumer-2007.03 tags: BMI operant conditioning review BCI date: 04-09-2007 14:25 gmt revision:0 [head]

PMID-17234696[0] Brain-computer interfaces: communication and restoration of movement in paralysis

  • A large gap between the promises of invasive animal and human BCI preparations and the clinical reality characterizes the literature: while intact monkeys learn to execute more or less complex upper limb movements with spike patterns from motor brain regions alone without concomitant peripheral motor activity usually after extensive training, clinical applications in human diseases such as amyotrophic lateral sclerosis and paralysis from stroke or spinal cord lesions show only limited success, with the exception of verbal communication in paralysed and locked-in patients.
  • attempts to train completely locked-in patients with BCI communication after entering the complete locked-in state with no remaining eye movement failed (!)
  • We propose that a lack of contingencies between goal directed thoughts and intentions may be at the heart of this problem. I'm not sure if 'contingencies' (something that can happen, but is generally not anticipated); should there not be a strong causal relationship between brain activity and prosthetic control?
  • still, the focus of this article are non-invasive BMIs.


hide / / print
ref: Wahnoun-2004.01 tags: BMI population_vector neural selection Brown 3D arizona ASU date: 04-06-2007 23:28 gmt revision:3 [2] [1] [0] [head]

PMID-17271333[0] Neuron selection and visual training for population vector based cortical control.

  • M1 and Pmd (not visual areas), bilateral.
  • a series of experiments designed to parameterize a cortical control algorithm without an animal having to move its arm.
  • a highly motivated animal observes as the computer drives a cursor move towards a set of targets once each in a center-out task.
    • how motivated? how did they do this? (primate working for its daily water rations)
  • I do not think this is the way to go. it is better to stimulate in the proper afferents and let the brain learn the control algorithm, the same as when a baby learns to crawl.
    • however, the method described here may be a good way to bootstrap., definitely.
  • want to generate an algorithm that 'tunes-up' control with a few tens of neurons, not hundreds as Miguel estimates.
  • estimate the tuning from 12 seconds of visual following (1.5 seconds per each of the 8 corners of a cube)
  • optimize over the subset of neurons (by dropping them) & computing the individual residual error.
  • their paper seems to be more of an analysis of this neuron-removal method.
  • neurons seem to maintain their tuning between visual following and brain-control.
  • they never actually did brain control

PMID-16705272[1] Selection and parameterization of cortical neurons for neuroprosthetic control

  • here they actually did neuroprosthetic control.
  • most units add noise to the control signal, a few actually improve it -> they emphasize cautious unit selection leaning to simpler computational/electrical systems.
  • point out that the idea of using chronically recorded neural signals has a very long history.. [2,3,4,5] [6] etc.
  • look like it took the monkeys about 1.6-1.8 seconds to reach the target.
    • minimum summed path length / distance to target = 3.5. is that good?


hide / / print
ref: Wood-2004.01 tags: spikes sorting BMI Black Donoghue prediction kalman date: 04-06-2007 21:57 gmt revision:2 [1] [0] [head]

PMID-17271178[0] automatic spike sorting for neural decoding

  • idea: select the number of units (and, indeed, clustering) based on the ability to predict a given variable. makes sense!
  • results:
    • human sorting: 13,5 cm^2 MSE
    • automatic spike sorting: 11.4 cm^2 MSE
      • yes, I know, the increase is totally dramatic.
  • they do not say if this could be implemented in realtime or not. hence, probably not.


hide / / print
ref: Sanchez-2004.01 tags: BMI nicolelis florida Carmena Principe date: 04-06-2007 21:02 gmt revision:3 [2] [1] [0] [head]

PMID-17271543[] http://hardm.ath.cx:88/pdf/sanchez2004.pdf


hide / / print
ref: Schwartz-2004.01 tags: Schwartz BMI prosthetics M1 review 2004 date: 04-05-2007 16:12 gmt revision:1 [0] [head]

PMID-15217341[0] Cortical Neuro Prosthetics

  • closed-loop control improves performance. see [1]
    • adaptive learning tech, when coupled to the adaptability of the cortex, suggests that these devices can function as control signals for motor prostheses.


hide / / print
ref: ONeill-1980.1 tags: BMI popular press Newsweek date: 04-04-2007 20:30 gmt revision:1 [0] [head]

I find this amusing - but appallingly old.


hide / / print
ref: Kamitani-2005.05 tags: BMI ATR fMRI date: 04-04-2007 15:20 gmt revision:0 [head]

PMID-15852014[] Decoding the visual and subjective contents of the human brain

  • used a linear SVM to decode visual orientation from voxel responses.
  • also were able to decode which direction of grating the subjects were paying attention to.


hide / / print
ref: Brockwell-2004.04 tags: particle_filter Brockwell BMI 2004 wiener filter population_vector MCMC date: 02-05-2007 18:54 gmt revision:1 [0] [head]

PMID-15010499[0] Recursive Bayesian Decoding of Motor Cortical Signals by Particle Filtering

  • It seems that particle filtering is 3-5 times more efficient / accurate than optimal linear control, and 7-10 times more efficient than the population vector method.
  • synthetic data: inhomogeneous poisson point process, 400 bins of 30ms width = 12 seconds, random walk model.
  • monkey data: 258 neurons recorded in independent experiments in the ventral premotor cortex. monkey performed a 3D center-out task followed by an ellipse tracing task.
  • Bayesian methods work optimally when their models/assumptions hold for the data being analyzed.
  • Bayesian filters in the past were computationally inefficient; particle filtering was developed as a method to address this problem.
  • tested the particle filter in a simulated study and a single-unit monkey recording ellipse-tracing experiment. (data from Rena and Schwartz 2003)
  • there is a lot of math in the latter half of the paper describing their results. The tracings look really good, and I guess this is from the quality of the single-unit recordings.
  • appendix details the 'innovative methodology ;)


hide / / print
ref: Marzullo-2006.12 tags: kipke BMI cingulate cortex 2006 date: 02-05-2007 17:14 gmt revision:1 [0] [head]

PMID-17190032[0] http://hardm.ath.cx:88/pdf/Marzullo2006_CingulateCortexBMI.pdf

  • motivation: ALS or PLS (primary lateral sclerosis) can damage upper motor neurons.
  • cingulate cortex has both cognitive and motor properties. & is involved in reward-based motor planning.
  • they give a long list of things that the cingulate cortex has been found to be involved in, including:
    • reward-based motor planning and reward expectancy
    • behavioral inhibition
    • stimulus-reward association
    • trace-conditioning
    • attention in complex discrimination tasks
    • error detection in humans
    • pain perception in human, too.
  • seven rats were able to modulate activity of neurons in cingulate cortex in order to recieve reward.
    • 52-84% percent of cingulate cortex neurons can be trained for a BMI; each seem to be independent.
  • michigan electrode, 16 channels.
  • auditory feedback.
  • food reward.
  • set the threshold based on the mean firing rate of SUA / MUA + a scalar times the stdev of the firing rate. the scalar was varied to allow 30-40% correct or operant rates.
  • used monte carlo simulations to verify the animal was performing above chance.
  • rat cortex is smooth :)
  • some cells increased their firing rate, some decreased (gaussian smoothed mean firing rate)
    • verified cell status with autocorrelogram.
result: cingulate cortex, like probably anywhere else, can come under voluntary control.


hide / / print
ref: Blankertz-2003.06 tags: BMI BCI EEG error classification motor commands Blankertz date: 0-0-2007 0:0 revision:0 [head]

PMID-12899253 Boosting bit rates and error detection for the classification of fast-paced motor commands based on single-trial EEG analysis

  • want to minimize subject training and maximize the major learning load on the computer.
  • task: predict the laterality of imminent left-right hand finger movements in a natural keyboard typing condition. they got ~15bits/minute (in one subject, ~50bits per minute!)
    • used non-oscilatory signals.
  • did a to detect 85% percent of error trials, and limited false-positives to ~2%

hide / / print
ref: Parra-2003.06 tags: BMI BCI EEG error correction ERN date: 0-0-2007 0:0 revision:0 [head]

PMID-12899266 Response error correction-a demonstration of improved human-machine performance using real-time EEG monitoring

  • the goal of an adaptive interface is to estimate variables correlated to human performance and adapt the HCI (human computer interface) = BCI accordingly.
    • use specific observable states to judge the subject's cognitive state, and use this information to adapt the BCI & maximize performance.
  • percieved errors are associated with a negative fronto-central deflection in the EEG signal = ERN, error-related negativity.
  • they can detect the ERN using a linear classifier within 100ms on a single-trial basis.
  • also have to remove eyeblink.

hide / / print
ref: Blankertz-2006.06 tags: BMI EEG ECoG competiton 2006 date: 0-0-2007 0:0 revision:0 [head]

PMID-16792282 http://hardm.ath.cx:88/pdf/BCIcompetition2006.pdf