One such researcher is Phil Kennedy, PhD.

Our immediate aims are to restore speech to mute and paralyzed individuals such as those with brainstem stroke or ALS (Amyotrophic lateral sclerosis, or Lou Gehrig's disease). One participant in the study, ER, was 16 years old when he was a passenger in a motor vehicle accident and suffered a brainstem stroke. Here is a picture of him with me during a recording session.

With ER, we did learn how to decode neural signals from his brain to recognize phonemes, words and phrases, among many other findings. We were not able to have him produce speech but we now know how to do the decoding and that is a very important result. I then had my speech cortex implanted for further data. The results from ER and I inspire us on to implant other locked-in persons to restore their speech. We need the funds to do these implants.

A. Personal Statement
My goal at this stage of my career is to develop a neural prosthetic such as a speech prosthesis that will allow mute individuals to speak in a near conversational manner. I am attacking the scientific and engineering challenges of such a prosthesis with an electrode that already provides longevity of signals. We need multiple electrodes to provide more channels for transmitting the signals. Following the completion of the speech prosthesis, I intend to pursue the activation of paralyzed muscles by cortical control signals, and then pursue the question of enhancement of the normal human brain. I aim to work with superb teams of neuroscientists, decoding experts and engineers to achieve these goals.

B. Positions and Honors

Working in Dr. Gadlage’s clinic Private practice 2017-date
Neural Signals Inc., CEO and Chief Scientist 1989-date
Community Neurological Clinic, Private Practice 1997-2016
Emory University, Dept. Neurology, Resident 1994-97
Emory University, Dept. Medicine, Resident 1993-94
Emory University, Yerkes Primate Center, Scientist 1988-97
Georgia Institute of Technology, Atlanta, GA
Research Scientist II, Bioengineering Center 1986-90
Senior Research Scientist, Director, Neuroscience. Lab 1990-97
Emory University, Dept. Physiology, Research Assoc. 1983-86
Northwestern University Graduate School, Chicago, IL
Graduate PhD Student, Dept. Physiology 1980-83
Univ. West. Ontario, London Ont., Canada
Post-Doctoral Fellow, Dept. Physiology 1978-80
Univ. West. Ontario Med. School, London, Ont., Canada
Resident, Dept. of Neurosurgery 1976-78
Resident in General Surgery, Dublin, Ireland 1972-76
Other Experience and Professional Memberships
Professional Affiliations: SFN; AAAS; AAN; AASM; IEEE Eng. Biol. & Med. Soc.; IBRO
Ad-hoc referee: NIH; NSF; Man in Motion, Rick Hensen Sp. Cord Res. Soc.; Easter Seal Res. Inst. Ontario; Sp. Cord Res. Foundation, TNRSE Associate editor,
Journal Referee: IEEE Trans.Biomed.Eng.; IEEE Trans.Rehab.Eng.; J Neurosci. Methods; Neurosci. Ltrs.; Prog. Brain Res., Frontiers, Nature Communications, Nature.

HONORS
Awards and media publications:
Wired Magazine 2016, MIT Technology Review 2016. Researcher of the Year, Neurotechnology Report, 2013; NPR 2016; Experimentarium Radio Interview Copenhagen (2010); Discover Sep 2008; Esquire Sep 2008, Scientific American Mind Oct 2008, Scientific American 2008, CNN (2007), PBS (2006), BBC (2006), Discovery Channel (2006), Scientist of the Year, Al Mann Foundation, 2004; Discover Magazine Award (1999), Atlanta Magazine’s Health Care Hero’s Award (1999), World Technology Network Award (London, England) (2000), Resource Forum Entrepreneur Award (2001), Life Time Member of National Registry of Who's Who. 1999; Pharmacia & Upjohn Award for outstanding research in CNS 1997; Parke-Davis Fellow 1994; National Research Council Senior Associate Award 1992; Alberta Heritage Medical Research Fellowship 1985; Northwestern University Graduate School Scholarship 1980-83; Muscular Dystrophy Association of Canada Fellowship 1979-80; Sheppard Memorial Prize, Ireland 1975;
US News and World Report, Chicago Tribune, USA Today, Boston Globe, Detroit News, Los Angeles Times, The Washington Post, New York Times, Newsday, San Francisco Chronicle, Atlanta Journal-Constitution, The Sciences, Current Science (Weekly Reader), New Scientist, Popular Science, Atlanta Magazine, CBS Evening News, CBS This Morning, ABC World News Tonight, ABC Good Morning America, Dateline NBC, The Learning Channel, Associated Press Television, Discovery (Canada), PBS “Frontiers of Medicine”, Tokyo Broadcasting System, Swedish Television, RTL German TV, Brazilian TV, FOX Network, CNN, BBC World News, Metro Source Radio (Syndicated Radio Broadcast), Associated Press Radio Network, National Public Radio “All Things Considered”, Forbes Magazine, Sunday Times, Creative Loafing.

C. Contributions to Science
1] Inferior Olivary Nucleus study in monkeys
In the late 70s I worked with Prof Vernon Brooks and George Ross at the University of Western Ontario, London, on studying the Inferior olivary nucleus (ION) in monkeys. The aim of the study was to elucidate the role of the ION in motor learning. The technique was to reversibly cool the ION by cooling. All cooling probes did not work at such a depth deep in the brainstem, so I developed a flexible probe made if Silastic tubing and a silver tip that selectively cooled the ION using nitrogen gas. We simultaneously recorded from the cerebellar Purkinje cells in the cerebellar cortex. Cooling produced a reduction of Purkinje cell firing of complex spikes but had no demonstrable effect on motor learning. Instead, the results showed that the trained monkeys developed slow oscillations of arm velocity profiles in a step tracking task, akin to the deficits in cerebellar lesions. Reversible cooling of the ION replicated lateral cerebellar lesions.
1. Movement programming depends on understanding of behavioral requirements. Brooks VB, Kennedy PR and Ross H-G. (1983) Physiol. & Behav. 31:561-563.
2. Participation of the Principal Olivary Nucleus in Neocerebellar Motor Control. Kennedy PR, Ross H-G and Brooks VB. Expt. BR Res. (1982) 47:95-104.
3. Chronic Implantation of Gas Cryoprobes in Monkey's Brainstem. Kennedy PR and Ross H-G. J. Neuroscience Methods (1980) 2:411-418.
4. X-ray Controlled Implantation of the Brain Stem. Kennedy P.R., H.-G. Ross. J. Neuroscience Methods (1980) 2:411-418.
2] Red Nucleus study in monkeys
Between 1980 and 1983, I completed my PhD dissertation with Prof. Jim Houk at Northwestern University in Chicago. I studied the magnocellular and parvocellular divisions using single unit recordings in monkeys with the aim of understanding their functional differences. I discovered the two subdivisions of the magnocellular division and confirmed their tight relationship to movement via the rubrospinal tract. The medium sized cells had not previously been described. These cells had low or no resting firing rates and lower firing rates during voluntary movements of the upper and lower extremities (and tails!), in contrast to the large neurons that had constant background firing and high firing rates during voluntary movements. In stark contrast, the small parvocellular neurons had variable resting firing rates and no appreciable relationship to voluntary movements. This led me to propose that the parvocellular neurons that project to the inferior olivary nucleus, (that projects to the cerebellar cortex and nuclei and thence back to the parvocellular nucleus and cerebral cortex) had an entirely different function, namely, they were part of a learning circuit. This hypothesis was published in Trends in Neuroscience.
1. Anatomic and functional contrast between Magnocellular and Parvocellular Red Nucleus. Kennedy PR, Houk JC and Gibson AR. Brain Research (1986), 364:124-136.
2. Parametric Relationships of Individual Digit Movements to Neuronal Discharges in Primate Magnocellular Red Nucleus. Kennedy PR. Brain Research (1987), 417:185-9.
3. Activity of primate Magnocellular Red Nucleus is related to hand and finger movements. Houk JC, Gibson AR, Harvey CF, Kennedy PR and Van Kan PLE. Behavioral Brain Research, 28 (1988) 201-206.
3] Parvocellular Red Nucleus study in rats
Between 1983 and 1986, I worked (a) Prof. Don Humphrey at Emory University, Atlanta again studying the red nucleus and ION, and Prof. Don Wigston, studying muscle regeneration in the Axolotl. I undertook a behavioral study in rats showing that parvocellular red nucleus (PRN) lesions do not lead to any deficit once the rat has recovered from a paralyzing lesion of the rubrospinal tract. In other words, if the PRN in intact when the rat is recovering from a lesion, the rat does not develop a further lesion when the PRN and Magnocellular Red nucleus is permanently lesioned. The conclusion being that the PRN is needed for recovery, but it is not needed for movement, or put another way, there is no motor deficit. (1). It was after these studies that I wrote the hypothesis (2).
The regeneration study with Prof. Wigston demonstrated that when regenerating spinal nerves into transposed muscles, the nerves found the appropriate muscles by varying their regenerating pathway (3). This allowed me to realize that regeneration is powerful. These studies lead to the development of the Neurtrophic Electrode.
1 The compensatory role of the parvocellular division of the red nucleus during reacquisition of coordinated motor tasks in operantly conditioned rats. Kennedy PR and Humphrey DR. Neuroscience Research (1987) 5:39-62.
2 Corticospinal, Rubrospinal and Rubro-olivary Projections: A unifying hypothesis. P.R. Kennedy, Trends in Neurosciences, 13(12):474-479 (1990).
3 Selective re-innervation of transplanted muscles by their original motoneurons in the axolotl. Kennedy PR and Wigston DJ. J. Neuroscience, (1987) 7(6):1857-1865.
4 Light labeling of red nucleus neurons following an injection of peroxidase conjugated wheat germ agglutinin into the inferior olivary nucleus of the rat. Kennedy PR. Neuroscience Letters (1987), 74:262-268.
4] Development of Chronic Recording Neurotrophic Electrode (NE)
In 1986, I opened a laboratory in the Biomedical Center in Georgia Tech, Atlanta GA. With the knowledge gained from my work outlined above, I realized that the only way to develop a long-term electrode was to grow the neuropil into the electrode, and not to insert the electrode into the neuropil because it will always lose its signal due to gliosis and micro-movements. Rat studies demonstrated long lasting signals when growth factors were used to entice neurites to grow into the glass cone tip with wires inside and retain their function. We used the barrel (whisker) cortex of rat as the test bed with signals enduring the lifetime of the rat (1).
These studies were continued at the Yerkes Primate Center of Emory University with neurosurgeon Dr. Bakay. The behaving monkey study demonstrated longevity of 14 months before destruction by the monkey. Furthermore, functional relationships to arm and hand movements were clearly demonstrated. Multiple signals were separated and differential relationships to flexion and extension clearly demonstrated (2). In addition, the histological appearance confirmed the histology as being similar to the rat histological appearance (3). Finally, the study demonstrated changes in single unit activity during task learning (4).
These results were presented to the FDA who gave permission for human studies.
1. A long-term electrode that records from neurites grown onto its recording surface. P.R. Kennedy, J. Neuroscience Methods, 29 (1989) 181-193.
2. Behavioral correlates of action potentials recorded chronically inside the Cone Electrode. P.R. Kennedy, R.A.E. Bakay and S.M. Sharpe. NeuroReport, 3:605-608, (1992).
3. The Cone Electrode: Ultrastructural Studies Following Long-Term Recording. P.R. Kennedy, S.Mirra and R.A.E. Bakay. Neuroscience Letters, 142:89-94, (1992).
4. Activity of single action potentials in monkey motor cortex during long-term task learning. Kennedy PR & Bakay RAE. Brain Research 760:251-4 (1997).
5] First human implantation of NE for communication via a computer
In 1986, Dr. Bakay and I implanted our first (MH) of six patients with a neurotrophic electrode at Emory University Hospital with the aim of demonstrating communication with a computer and thereby allowing her to communicate. MH (49 years old) was a severely locked-in patient with multiple medical problems from which she shortly succumbed. The NE was implanted in the arm area as confirmed by functional MRI study. This study demonstrated that she could voluntarily control the signal in a binary fashion (1). This confirmed the monkey studies.
Following subject MH, Dr. Bakay and I implanted JR in his arm area in 1998. Over four years of recording, JR provided multiple data points before he succumbed to his underlying medical problems. He demonstrated control of individual single units separated from the multi-unit data stream so that he could spell out words on the monitor in two dimensions. He was able to spell spontaneously (2). He clearly demonstrated another first: The emergence of cursor related cortex after weeks of training the signals that arose from the implanted arm area (3).
Two subjects were implanted with nothing more than a skull screw to augment the EEG signal as happens with breach rhythm. They demonstrated control of a switch (4).
Finally an ALS subject with digit micro-movement was implanted in the arm area. EMG activity and digit micro-movements were correlated with the single units and local field potentials (5). The aim here was to see if we could progress to controlling movements.
1. Restoration of neural output from a paralyzed patient using a direct brain connection. Kennedy P.R. and R.A.E.Bakay. NeuroReport 9,1707-11, 1998.
2. Direct control of a computer from the human central nervous system. Kennedy PR, Bakay RAE, Adams K, Goldthwaite J, and M. Moore. IEEE Trans. Rehab. Eng., 8(2), 198-202, 2000.
3. Dynamic interplay of neural signals during the emergence of cursor related cursor in a human implanted with the Neurotrophic electrode. Kennedy PR and King B. CH 7 in Neural Prostheses for Restoration of Sensory and Motor Function. Eds. Chapin J and Moxon, K. CRC Press, 2001.
4. Using Human Extra-cortical Local Field Potentials to Control a Switch. Kennedy PR, Dinal Andreasen, Princewill Ehirim, Brandon King, Todd Kirby, Hui Mao, Melody Moore. J. Neural Eng. 1: 72-77, 2004.
5. Correlations between human motor cortical local field potentials, action potentials, contralateral arm EMG activity and digit movements. Kennedy PR, Dinal Andeasen, Brandon King, Todd Kirby, Hui Mao, Melody Moore, Princewill Ehirim. J. Neural Engineering 2005.
6] First human implantation of NE for Speech prosthesis
Following a lot of heart searching, I felt it unethical to perform craniotomies when these subjects had some muscle movements. Instead we developed an EMG recording device (the Impulse) that wirelessly transmitted its signals to a computer to allow switch control of the computer. The unique role of the implantable NE would be to control movement or generate near-conversational speech. So subject ER’s speech cortex was implanted in 2004 with signals continuing until 2015 when he became too ill to even sit up. His two electrodes produced over 30 individual single units that allowed identification of half the English phonemes. With colleagues Frank Gunther and Jonathan Brumberg from MIT and Boston University, he demonstrated control of the cursor such that he could move from one vowel phoneme to another (1,2&3). In addition, we demonstrated that changes in emotional state affect firing activity of units (4).
Following subject ER’s slow decade of recording, we realized that it is essential to record from speaking subjects who can produce overt and covert speech and compare signals. I became that subject and was implanted in 2014 by Dr. Cervantes. The analysis to date shows a total of 65 single units from three electrodes, identification of 32 phonemes (of 39 total), sensory (light touch and pin) and motor relationships, and an ongoing analysis of single unit bursting patterns (5,6). Of importance, the data demonstrate that 12 – 20 Hz beta peaks are present at the onset, inflections points and offset of speech. We are prioritizing these data now (see next section).
1.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. 2009 Dec 9;4(12):e8218.
2. Brumberg J., Wright EJ, Andersen D, Guenther FH and Kennedy PR. Classification of intended phoneme production from chronic intracortical microelectrode recordings in speech motor cortex. Frontiers in Neuroscience 5(65)1-14, 2011.
3. Brumberg JS, Nieto-Castanon A, Kennedy PR, Guenther FH. Brain-Computer Interfaces for Speech Communication. Speech Commun. 2010 Apr 1;52(4):367-379.
4. Kennedy, PR. Changes in emotional state modulate neuronal firing rates of human speech motor cortex: A case study in long-term recording. Neurocase 2011, 17(5), 381-393.
5. Kennedy P. R., C. Gambrell and N. Shih, Detection of phonemes, short words and phrases from single units and 12-20 Hz frequency Beta band data during overt and covert speech recorded chronically from a speaking human. SFN Abstracts 2016.
6. Kennedy P.R., Andre Cervantes. Modulation of single unit firing rates during phoneme production from the speech area of an intact human. SFN Abstracts 2015.
7. Kennedy P.R., Andreasen D.S., Bartels J., Ehirim P., E. Joe Wright, E.J., Seibert, S., Cervantes, A.J. Validation of Neurotrophic Electrode long term recordings in Human Cortex. Accepted by Handbook of Biomedical Engineering 2017.
8. Kennedy P.R., Gambrell C, Ehirim P, and Cervantes A. Advances in the development of a speech prosthesis. Book chapter in Brain-Machine Interfaces: Uses and Developments accepted 2017.
Gearin, M. and Kennedy, P.R. 2020. Histological confirmation of myelinated neural filaments within the tip of the Neurotrophic Electrode after a decade of neural recordings. Frontiers in Human Neuroscience 21 April 2020 https://doi.org/10.3389/fnhum.2020.00111

D. Additional Information: Research Support and/or Scholastic Performance

INVITED PRESENTATIONS

2021 Zoom talk at Univ. Newcastle, England: Seven criteria for a speech prosthetic.
2020 Brain computer panel discussion
2017 Panel discussion of future developments in AI and robots. Summit Knowledge conference. Dubai, UAE, Nov 22.
2016 Talk on Speech prosthesis seminar. Brown Univ., Rhode Island, USA
2013 Steps in the development of a speech prosthesis. Grand Rounds, Univ. of Oxford, England, Dept. of Neurology and Neurosurgery due Feb 8.
2013 Developing a speech prosthesis. Lecture, Univ. Newcastle, Dept. Neuroscience Feb 6
2013 Electrode design and implementation. Lecture, University College Cork, Ireland Feb 4
2012 Patent issues in the development of Speech Prosthesis, University College London.
2012 Development of Speech Prosthesis, Oxford Dept. Neurosurgery.
2012 Engineering the Speech Prosthesis, Tyndall National labs, Cork Ireland.
2012 Neural Prostheses Meeting, Univ. Utah. June 17th to 20th
2012 American Academy of Neurology, New Orleans, April 24
2011 Univ. of Utrecht, talk, May 20 to 21st, Utrecht The Netherlands.
2010 Building Better Brains: Neural Prosthetics and Beyond. Given Inst., Aspen. NYAS Sept
2010 Brain Machine Interfaces - Implications for science, clinic and society, Lund Univ. Aug. Etc.

INTELLECTUAL PROPERTY

ISSUED:

“Implantable Neural Electrode.” Patent #: 4,852,573, issued on August 1, 1989.
“System and Method for Speech Generation from Brain Activity.” Serial number: 1/007,380. Filing date: 12/08/2003. Issued September 2007.
Apparatus and method for detecting neural signals and using neural signals to drive external functions. Issued March 6, 2007. US: 7,187,967.B2
Medication Dispensing Device. Filed 5/29/2009. Issued August 30, 2011. US 8,009,040.B2
“Quantum Dot Neurotrophic Electrode Arrays”. Filed 3/6/ 2007. Serial Number 60/893,16. Foreign (PTC) filing: Neurotrophic Electrode Neural Interface Employing Quantum Dots.” Filing date: 3/5/2008. SSerial Number 12042742

PENDING:

“Neural Electrode Array.” Serial #: 11/096,897. Filing date: 04/01/2005.
“Software controlled electromyogram control system”. Serial #: 10/881,923. Filing date: 06/30/2004.
“Detecting Neural Signals and using same to drive external functions.” Serial #: 10/675,703. Filing date: 9/30/2003.
"Speech Prosthesis Employing Beta Peak Detection"; Serial number: 15/800,589 Filed November 2017.
“Wireless powering of medical devices”.