Accelerating Translational Neurotechnology: 4th Annual Aspen Brain Forum

Resources

Bionic skeletons and beyond

Courtine G, Song B, Roy RR, et al. Recovery of supraspinal control of stepping via indirect propriospinal relay connections after spinal cord injury. Nat Med. 2008;14(1):69-74.

Dominici N, Keller U, Vallery H, et al. Versatile robotic interface to evaluate, enable and train locomotion and balance after neuromotor disorders. Nat Med. 2012;18(7):1142-7.

Van den Brand R, Heutschi J, et al. Restoring voluntary control of locomotion after paralyzing spinal cord injury. Science. 2012;336(6085):1182-5.

Bridging Bionics Foundation
The Bridging Bionics Foundation provides funding, supports education, and advances research and development for exoskeletons and bionic technology.

Ekso Bionics
Ekso Bionics builds pioneering robotic exoskeletons to augment human strength, endurance, and mobility.

Keynote panel: the economic potential of neurotechnology—funding landscape and economic impact

NDI Medical
NDI is a hybrid venture capital and commercialization firm focusing exclusively on innovative neurodevice technologies to restore lost neurological function, prevent damage, and reduce the painful effects of disease and injury.

Neurotechnology Industry Organization
NIO is the trade association representing companies involved in neuroscience—drugs, devices, and diagnostics companies; brain research centers; and advocacy groups.

NeuroVigil Inc.
NeuroVigil is using brain-wave data obtained by merging neuroscience, non-invasive wireless brain-recording technology, and advanced computational algorithms to assist with diagnosis and treatment.

QiG Group
QiG Group is focused on promoting the creative design, development, and speed to market of new medical devices in the therapy fields of cardiovascular health, neuromodulation, and orthopedics.

Neuroprosthetics and neuroengineering

Brockwell AE, Kass RE, Schwartz AB. Statistical signal processing and the motor cortex. Proc IEEE. 2007;95:881-98.

Chase SM, Kass RE, Schwartz AB. Behavioral and neural correlates of visuomotor adaptation observed through a brain-computer interface in primary motor cortex. J Neurophysiol. 2012;108(2):624-44.

Chase SM, Schwartz AB. Inference from populations: going beyond models. Prog Brain Res. 2011;192:103-12.

Chase SM, Schwartz AB, Kass RE. Bias, optimal linear estimation, and the differences between open-loop simulation and closed-loop performance of spiking-based brain-computer interface algorithms. Neural Netw. 2009;22(9):1203-13.

Chase SM, Schwartz AB, Kass RE. Latent inputs improve estimates of neural encoding in motor cortex. J Neurosci. 2010;30(41):13873-82.

Courtine G, Song B, Roy RR, et al. Recovery of supraspinal control of stepping via indirect propriospinal relay connections after spinal cord injury. Nat Med. 2008;14(1):69-74.

Dominici N, Keller U, Vallery H, et al. Versatile robotic interface to evaluate, enable and train locomotion and balance after neuromotor disorders. Nat Med. 2012;18(7):1142-7.

Fisher R, Salanova V, Witt T, et al. Electrical stimulation of the anterior nucleus of thalamus for treatment of refractory epilepsy. Epilepsia. 2010;51(5):899-908.

Morrell MJ, RNS System in Epilepsy Study Group. Responsive cortical stimulation for the treatment of medically intractable partial epilepsy. Neurology. 2011;77(13):1295-304.

Schwartz AB. Useful signals from motor cortex. J Physiology. 2007;579(Pt 3):581-601.

Schwartz AB, Cui XT, Weber DJ, Moran DW. Brain controlled interfaces: movement restoration with neural prosthetics. Neuron. 2006;52(1):205-20.

Stead M, Bower M, Brinkmann BH, et al. Microseizures and the spatiotemporal scales of human partial epilepsy. Brain. 2010;133(9):2789-97.

Van den Brand R, Heutschi J, Barraud Q, et al. Restoring voluntary control of locomotion after paralyzing spinal cord injury. Science. 2012;336(6085):1182-5.

Van Hemmen JL, Schwartz AB. Population vector code: a geometric universal as actuator. Biol Cybern. 2008;98:509-18.

Velliste M, Perel S, Spalding MC, et al. Cortical control of a prosthetic arm for self-feeding. Nature. 2008;453:1098-101.

Viventi J, Kim DH, Vigeland L, et al. Flexible, foldable, actively multiplexed, high-density electrode array for mapping brain activity in vivo. Nat Neurosci. 2011;14(12):1599-605.

Worrell GA, Parish L, Cranstoun SD, et al. High-frequency oscillations and seizure generation in neocortical epilepsy. Brain. 2004;127(Pt 7):1496-506.

Neuroimaging

Apostolova LG, Green AE, Babakchanian S, et al. Hippocampal atrophy and ventricular enlargement in normal aging, mild cognitive impairment (MCI), and Alzheimer Disease. Alzheimer Dis Assoc Disord. 2012;26(1):17-27.

Braskie Meredith N, Toga Arthur W, Thompson PM. Recent advances in imaging Alzheimer's disease. J Alzheimers Dis. 2013;33 Suppl 1(1):S313-27.

Carbon M, Reetz K, Ghilardi MF, et al. Early Parkinson's disease: longitudinal changes in brain activity during sequence learning. Neurobiol Dis. 2010;37(2)455-60.

Herrup K, Carrillo MC, Schenk D, et al. Beyond amyloid: getting real about nonamyloid targets in Alzheimer's disease. Alzheimers Dement. 2013;9(4):452-8.

Hirano S, Shinotoh H, Eidelberg D. Functional brain imaging of cognitive dysfunction in Parkinson's disease. J Neurol Neurosurg Psychiatry. 2012;83(10):963-9.

Ko JH, Mure H, Tang CC, et al. Parkinson's disease: increased motor network activity in the absence of movement. J Neurosci. 2013;33(10):4540-9.

Mosconi L, Berti V, Glodzik L, et al. Pre-clinical detection of Alzheimer's disease using FDG-PET, with or without amyloid imaging. J Alzheimers Dis. 2010;20(3):843-54.

Mosconi L, Mistur R, Switalski R, et al. FDG-PET changes in brain glucose metabolism from normal cognition to pathologically verified Alzheimer's disease. Eur J Nucl Med Mol Imaging. 2009;36(5):811-22.

Spetsieris P, Ma Y, Peng S, et al. Identification of disease-related spatial covariance patterns using neuroimaging data. J Vis Exp. 2013;(76).

Tang C, Poston K, Dhawan V, Eidelberg D. Abnormalities in metabolic network activity precede the onset of motor symptoms in Parkinson's disease. J Neurosci. 2010;30(3):1049-56.

Toga AW, Dinov ID, Thompson PM, et al. The Center for Computational Biology: resources, achievements, and challenges. J Am Med Inform Assoc. 2012;19(2):202-6.

Toga AW, Thompson PM. Connectomics sheds new light on Alzheimer's disease. Biol psychiatry. 2013;73(5):390-2.

Neuromodulation, optogenetics, and deep brain stimulation

Agnesi F, Johnson MD, Vitek JL. Deep brain stimulation: how does it work? Handb Clin Neurol. 2013;116:39-54.

Bernstein JG, Garrity PA, Boyden ES. Optogenetics and thermogenetics: technologies for controlling the activity of targeted cells within intact neural circuits. Curr Opin Neurobiol. 2012;22(1):61-71.

Chow BY, Boyden ES. Optogenetics and translational medicine. Sci Transl Med. 2013;5(177):177ps5.

Chow BY, Han X, Boyden ES. Genetically encoded molecular tools for light-driven silencing of targeted neurons. Prog Brain Res. 2012;196:49-61.

De Hemptinne C, Ryapolova-Webb ES, Air EL, et al. Exaggerated phase-amplitude coupling in the primary motor cortex in Parkinson disease. Proc Natl Acad Sci U S A. 2013;110(12):4780-5.

Gradinaru V, Mogri M, Thompson KR, et al. Optical deconstruction of parkinsonian neural circuitry. Science. 2009;324(5925):354-9.

Hendrix CM, Vitek JL. Toward a network model of dystonia. Ann NY Acad Sci. 2012;1265:46-55.

Holtzheimer PE, Kelley ME, Gross RE, et al. Subcallosal cingulate deep brain stimulation for treatment-resistant unipolar and bipolar depression. Arch Gen Psychiatry. 2012;69(2):150-8.

Johnson MD, Zhang J, Ghosh D, et al. Neural targets for relieving parkinsonian rigidity and bradykinesia with pallidal deep brain stimulation. J Neurophysiol. 2012;108(2):567-77.

Kennedy SH, Giacobbe P, Rizvi SJ, et al. Deep brain stimulation for treatment-resistant depression: follow-up after 3 to 6 years. Am J Psychiatry. 2011;168(5):502-10.

Kubu CS, Malone DA, Chelune G, et al. Neuropsychological outcome after deep brain stimulation in the ventral capsule/ventral striatum for highly refractory obsessive-compulsive disorder or major depression. Stereotact Funct Neurosurg. 2013;91(6):374-8.

Lozano AM, Giacobbe P, Hamani C, et al. A multicenter pilot study of subcallosal cingulate area deep brain stimulation for treatment-resistant depression. J Neurosurg. 2012;116(2):315-22.

Malone DA Jr., Dougherty DD, Rezai AR, et al. Deep brain stimulation of the ventral capsule/ventral striatum for treatment-resistant depression. Biol Psychiatry. 2009;65(4):267-75.

Mayberg HS, Lozano AM, Voon V, et al. Deep brain stimulation for treatment-resistant depression. Neuron. 2005;45(5):651-60.

Miocinovic S, Somayajula S, Chitnis S, Vitek JL. History, applications, and mechanisms of deep brain stimulation. JAMA Neurol. 2013;70(2):163-71.

Okun MS, Gallo BV, Mandybur G, et al. Subthalamic deep brain stimulation with a constant-current device in Parkinson's disease: an open-label randomised controlled trial. Lancet Neurol. 2012;11(2):140-9.

Riva-Posse P, Holtzheimer PE, Garlow SJ, Mayberg HS. Practical considerations in the development and refinement of subcallosal cingulate white matter deep brain stimulation for treatment-resistant depression. World Neurosurg. 2013;80(3-4):S27.e25-34.

Whitmer D, de Solages C, Hill B, et al. High frequency deep brain stimulation attenuates subthalamic and cortical rhythms in Parkinson's disease. Front Hum Neurosci. 2012;6:155.

Stem cells and therapies

Cooper O, Hallett P, Isacson O. Using stem cells and iPS cells to discover new treatments for Parkinson's disease. Parkinsonism Relat Disord. 2012;18 Suppl 1:S14-6.

Cooper O, Seo H, Andrabi S, et al. Pharmacological rescue of mitochondrial deficits in iPSC-derived neural cells from patients with familial Parkinson's disease. Sci Transl Med. 2012;4(141):141ra90.

McLeod M, Hong M, Sen A, et al. Transplantation of bioreactor-produced neural stem cells into the rodent brain. Cell Transplant. 2006;15(8-9):689-97.

Mukhida K, Baker KA, Sadi D, Mendez I. Enhancement of sensorimotor behavioral recovery in hemiparkinsonian rats with intrastriatal, intranigral, and intrasubthalamic nucleus dopaminergic transplants. J Neurosci. 2001;21(10):3521-30.

Mukhida K, Hong M, Miles GB, et al. A multitarget basal ganglia dopaminergic and GABAergic transplantation strategy enhances behavioural recovery in parkinsonian rats. Brain. 2008;131(Pt 8):2106-26.

Mukhida K, Mendez I, McLeod M, et al. Spinal GABAergic transplants attenuate mechanical allodynia in a rat model of neuropathic pain. Stem Cells. 2007;25(11):2874-85.

Sundberg M, Bogetofte H, Lawson T, et al. Improved cell therapy protocols for Parkinson's disease based on differentiation efficiency and safety of hESC-, hiPSC-, and non-human primate iPSC-derived dopaminergic neurons. Stem Cells. 2013;31(8):1548-62.

Keynote address: a retinal prosthesis—from idea to clinical application

Ahuja AK, Dorn JD, Caspi A, et al. Blind subjects implanted with the Argus II retinal prosthesis are able to improve performance in a spatial-motor task. Br J Ophthalmol. 2011;95(4):539-43.

Barry MP, Dagnelie G, Argus II Study Group. Use of the Argus II retinal prosthesis to improve visual guidance of fine hand movements. Invest Ophthalmol Vis Sci. 2012;53(9):5095-101.

Da Cruz L, Coley BF, Dorn J, et al. The Argus II epiretinal prosthesis system allows letter and word reading and long-term function in patients with profound vision loss. Br J Ophthalmol. 2013;97(5):632-6.

Humayun MS, Dorn JD, da Cruz L, et al. Interim results from the international trial of Second Sight's visual prosthesis. Ophthalmology. 2012;119(4):779-88.

Keynote panel: accelerating neurotechnology research—a global perspective

BIRD Foundation
The BIRD Foundation was established by the U.S. and Israeli governments in 1977 to generate mutually beneficial cooperation between the private sectors of the U.S. and Israeli high-tech industries, including in communications, life sciences, electronics, electro-optics, software, and other areas.

(BRAIN) Initiative
The NIH Brain Research through Advancing Innovative Neurotechnologies (BRAIN) Initiative is part of a new Presidential focus aimed at revolutionizing our understanding of the human brain.

DARPA. Defense Sciences Office. Neuroscience.
DSO develops and leverages neurophysiological sensors, neuroimaging, cognitive science, and molecular biology to provide support, protection, and tactical advantage to warfighters who perform under the most challenging operational conditions.

Human Brain Project
The goal of the Human Brain Project is to build a completely new information computing technology infrastructure for neuroscience and for brain-related research in medicine and computing, catalyzing a global collaborative effort to understand the human brain and its diseases and ultimately to emulate its computational capabilities.

Innovative Medicines Initiative
The Innovative Medicines Initiative (IMI) is Europe's largest public–private initiative. It aims to speed the development of better and safer medicines for patients.

State of the disease lectures

Chapin H, Bagarinao E, Mackey S. Real-time fMRI applied to pain management. Neurosci Lett. 2012;520(2):174-81.

Eidelberg D, Martin W. Different β-amyloid binding patterns in Alzheimer and Parkinson diseases: it's the network! Neurology. 2013;81(6):516-7.

Gavett BE, Cantu RC, Shenton M, et al. Clinical appraisal of chronic traumatic encephalopathy: current perspectives and future directions. Curr Opin Neurol. 2011;24(6):525-31.

Holtzheimer PE, Kelley ME, Gross RE, et al. Subcallosal cingulate deep brain stimulation for treatment-resistant unipolar and bipolar depression. Arch Gen Psychiatry. 2012;69(2):150-8.

Holtzman DM, Morris JC, Goate AM. Alzheimer's disease: the challenge of the second century. Sci Transl Med. 2011;3(77):77sr1.

Regulatory and ethical challenges in translating neuroscience research

Azemi E, Lagenaur CF, Cui XT. The surface immobilization of the neural adhesion molecule L1 on neural probes and its effect on neuronal density and gliosis at the probe/tissue interface. Biomaterials. 2011;32(3):681-92.

FDA. Medical Devices.
Cutting-edge innovations in neurotechnology from submitted abstracts

Hanson TL, Fuller AM, Lebedev MA, et al. Subcortical neuronal ensembles: an analysis of motor task association, tremor, oscillations, and synchrony in human patients. J Neurosci. 2012;32(25):8620-32.

Kolarcik CL, Bourbeau D, Azemi E, et al. In vivo effects of L1 coating on inflammation and neuronal health at the electrode-tissue interface in rat spinal cord and dorsal root ganglion. Acta Biomaterialia. 2012;8(10):3561-75.

Nie K, Ling L, Bierer SM, et al. An experimental vestibular neural prosthesis: design and preliminary results with rhesus monkeys stimulated with modulated pulses. IEEE Trans Biomed Eng. 2013;60(6):1685-92.

Phillips C, Defrancisci C, Ling L, et al. Postural responses to electrical stimulation of the vestibular end organs in human subjects. Exp Brain Res. 2013;229(2):181-95.

Swan BD, Grill WM, Turner DA. Investigation of deep brain stimulation mechanisms during implantable pulse generator replacement surgery. Neuromodulation. 2013. [Epub ahead of print]

Welle C, Krauthamer V. FDA regulation of invasive neural recording electrodes: a daunting task for medical innovators. IEEE Pulse. 2012;3(2):37-41.

Sponsors

Silver Sponsors

Academy Friends

Craig H. Neilsen Foundation

Ripple

Presented by

Neurotechnology is moving at a rapid pace. Speakers at the conference presented advances in treating neurological diseases and injuries, ranging from stem cell-based therapies to improve function in neurodegenerative disorders to neuroprosthetic skeletons to help restore movement to patients with paralysis, FDA-approved devices to recover some sight in retinitis pigmentosis-induced blindness, and devices in development to treat Meniere's disease-associated vertigo.

However, the brain is very complex and our knowledge of neuroscience remains significantly limited. Progress in neurotechnology will require continued investigation of basic neuroscience, novel advances in imaging, and new biomarkers to enable earlier identification of patients. In particular, a whole-brain approach to treatment will require a better understanding of how neural circuits interact. Because of the variability in brain architecture among different patients and the complexity of the brain itself, individualized treatment requires massive amounts of data, systems to analyze these data, and collaborations to utilize data effectively.

Collaboration was a key theme for this year's conference, particularly between scientists and developers but also including manufacturers and regulatory agencies. Large-scale initiatives, including the Obama administration's BRAIN Initiative (Brain Research through Advancing Innovative Neurotechnologies) and the European Commission's Human Brain Project are driving collaborations that will help not only to spur the development of new technologies but also to bring these technologies to market.

It is important to understanding how to market new technologies and to identify barriers to success early in development. Speakers emphasized that knowing how to raise capital and approach the regulatory process is as important as the scientific development of a new product. There are business and scientific concerns as well as ethical considerations, and developers must ensure that they address these barriers early in the development process to help ensure success.

Many speakers expressed excitement about current work in neurotechnology. The conference featured moving videos demonstrating neurotechnology at work, showing severely depressed patients smiling after deep brain stimulation, Parkinson's disease patients confidently walking, completely paralyzed rats walking again, completely paralyzed people moving robotic arms with their thoughts, and blind people reading characters on a computer screen. In the opening session, Amanda Boxtel, who is paralyzed from the waist down, walked around the conference room using a bionic exoskeleton. Many new technologies are still in their infancies, providing hope that some of the most difficult and debilitating neurological diseases and injuries will have better treatment options in the future.

Moderator:
Orla M. Smith, Science Translational Medicine
 
Speakers:
Grégoire Courtine, Swiss Federal Institute of Technology (EPFL), Switzerland
Nathan Harding, Ekso Bionics
Amanda Boxtel, Bridging Bionics Foundation

Highlights

• In animal models, neurotechnology advances have restored movement potential to animals with spinal cord lesions.
• Bionic exoskeletons that allow for movement are now available for rehabilitation of people with lost motor function.

Reanimating the spinal brain

Grégoire Courtine and his colleagues at the Swiss Federal Institute of Technology (EPFL) are working to restore function to completely paralyzed patients. Using an electrochemical neuroprosthesis—an implanted device that delivers bilateral electrical stimuli and a cocktail of pharmaceuticals—combined with a robotic system that provides support and freedom of movement, Courtine's team has been able to restore hind leg movement in rats that lost this function through spinal cord lesions. Although the device successfully restores motor control in animal models, the rats do not recover balance, and this deficiency needs to be corrected. The researchers plan to try a similar approach in humans, but Courtine noted that it has been difficult to obtain regulatory approval to initiate studies.

How can machines help: bionic exoskeletons

Amanda Boxtel in the Ekso exoskeleton. (Image courtesy of Nathan Harding)

Ekso Bionics is pursuing a complementary approach, developing bionic exoskeletons for use in rehabilitation and to provide functional improvements. Nathan Harding, CEO of Ekso Bionics, explained that the exoskeleton is an old idea that has advanced rapidly in the past decade as technological improvements allowed for power consumption to drop by three orders of magnitude. The company's exoskeleton is designed to assist in rehabilitation, and more than 40 devices are currently in use at rehabilitation centers in the U.S. The exoskeleton allows users to bear their full body weight on their legs, potentially improving bone density. Users can stand with their joints aligned and are not confined to taking steps on a treadmill but can experience sensory movement through space. Harding believes that the production of brain-to-machine devices is inevitable but at least a decade off.

Amanda Boxtel of the Bridging Bionics Foundation demonstrated wearing an Ekso exoskeleton on stage. Boxtel lost motor function from the waist down two decades ago, at age 24, in a skiing accident. She retains some minor hip flexor function, which enables her to operate the Ekso exoskeleton. She uses the exoskeleton to assist in gait-training rehabilitative therapy, and she believes that the device helps the brain to rewire neural pathways. It has power settings that allows her to use it for exercise, which significantly improves her quality of life. As she walked through the auditorium, Boxtel remarked "I love the role reversal—you are all sitting down and I am walking."

Moderator:
Zack Lynch, Neurotechnology Industry Organization
 
Panelists:
Philip Low, NeuroVigil Inc.
Geoffrey B. Thrope, NDI Medical
Alan Mock, QiG Group

Highlights

• Brain illnesses represent significant morbidity worldwide and have a massive economic impact.
• Neurotechnology advances need to be commercially viable in order to have an impact once they reach the market.

Funding landscape and economic impact of neurotechnology

There are an estimated 2 billion cases of brain illness worldwide, with 100 million in the U.S. alone. These illnesses include addiction, anxiety, Alzheimer's disease (AD), chronic pain, migraine, and many others. Their overall economic burden is enormous, standing at around $1 trillion dollars annually in the U.S. and spurring much effort to develop neurotechnologies. The number of venture capital deals in neurotechnology skyrocketed in 2012, but many of these deals represent small, early-stage investments. Zack Lynch of the Neurotechnology Industry Organization led a panel to discuss the challenges that exist for companies investing in neurotechnology.

Panelists Philip Low of NeuroVigil Inc., Geoffrey B. Thrope of NDI Medical, and Alan Mock of QiG Group all have significant experience in developing nascent neurotechnology. They discussed intellectual property, how to raise capital, and how to bring neurotechnology to market. The panelists agreed that patents need to be legally defensible. They also noted that universities and academics frequently overvalue their patents during technology transfer. Mock asserted that "most patents are good ideas with no commercial value"; but patents must hold some commercial value to be marketable. Such value includes ensuring that when a technology reaches the market, reimbursement channels are available (that is, people will be able to buy the product). Many products never receive funding and do not make it to market. Companies investing in neurotechnology may shy away from breakthrough technologies because of their low chance of commercial success. Reimbursement is complicated by the complexity in international markets. Cost-effectiveness data and a compelling value proposition can be very useful in surmounting these commercial barriers.

According to Thrope, it takes over $100 million to bring a technology to market. There are several avenues to generate funding for new technologies. State and federal grants can be used to advance and validate the science behind the technology. This validation can later be leveraged for venture capital; however, the panelists acknowledged that working with venture capital companies can be challenging and can result in loss of control over the project. Thrope commented that as a new technology grows and comes closer to market, some control over its development is always lost to investors or partners.

Low described a different avenue to raise capital; he believes that the best funding comes from revenue. By partnering with Roche, his company was able to generate revenue and provide market validation for new products. The panelists agreed that bringing a technology to market always costs more than is expected, and fundraising is perpetual. Therefore, people who have gone through the process before can provide invaluable resources to increase the likelihood of success.

Brain-related disorders in the U.S. and worldwide. (Image courtesy of Zack Lynch)

Chair:
Andrew Schwartz, University of Pittsburgh
 
Speakers:
Grégoire Courtine, Swiss Federal Institute of Technology (EPFL), Switzerland
Steven M. Chase, Carnegie Mellon University
Brian Litt, University of Pennsylvania

Highlights

• It is now possible to decode neuronal impulses in the brain for intended movement and to translate these data into robotic limb movement. However, there is still much work to be done to optimize the process.
• Larger-scale decoding of neuronal networks and improved lead technologies have the potential to significantly improve neuroprosthetic function and disease-targeting technological interventions.

Neuroprosthetic technologies to restore motor function after severe paralysis

In the session that focused on neuroprosthetics and neuroengineering, Grégoire Courtine from EPFL spoke about "the spinal brain." His group has developed a neurorehabilitation multisystem that combines chemical, electrical, and robotic signaling to restore neural function. Targeting spinal cord trauma (SCT) in a rodent model, the team replaced neuronal excitation and modulation with drug and electrical inputs in the spinal column. Electrochemical stimuli—stretchable and flexible electrode arrays that conduct current—produce monosynaptic and polysynaptic responses, allowing researchers to map motor coordination. With the addition of a multidirectional robotic interface, the team could stimulate extension and flexion of muscles and produce a linear increase in overall muscle activity. The robotic interface allows for adjustment, achieves precision movement through a feedback loop, and creates four dimensions of freedom.

The researchers are now grappling with how to build upon this success in animal models to bring a multifaceted approach to humans. Courtine noted that nonhuman primate models are only in the development stages. Obtaining ethical approval for human studies has been challenging, but the group has developed a massive robotic interface that allows for three dimensions of freedom and is designed to be used in human trials.

Restoring movement in paralyzed rats using chemical, electrical, and robotic systems. (Image courtesy of Grégoire Courtine)

Accessing and decoding neuronal activity for human–machine interfaces

Andrew Schwartz from the University of Pittsburgh discussed efforts to access and record neural activity and translate the data into a format that can be used to control external devices. His team focuses on prosthetic devices such as a robotic limbs, robotic hands, and even a cursor on a computer screen.

To build neural prosthetics we need to decode neural signaling, to create a map that describes the linkage between neuronal firing rates and the movement of the prosthesis. Three decades ago, a landmark experiment by Georgopoulos recorded the firing rates of neurons in the primary motor cortex of monkeys. He discovered that the rate of firing differed according to the direction of the movement in the monkeys' limbs: the firing rate of a neuron is highest in the cell's "preferred direction" and falls off as a function of the angle between the movement and this direction. This discovery prompted the development of the first decoding algorithm, the population vector algorithm, which scientists use to predict the direction of movements based on the number of times a particular neuron fires. In this model, the more a neuron fires, the more it "pushes" the movement towards its own preferred direction. Assuming that the length of that population vector is the desired speed, it is possible to also decode the velocity of the movement.

Schwartz's research is focused on mapping intended movement on an instant-by-instant basis. He noted that classical biology advances our understanding of a system by reducing its variables and components to the smallest possible scale so that they can be extensively studied and assembled into a larger framework. Neuroscience may require an opposite approach, in which scientists work to derive meaningful relationship by investigating many components of a system. Schwartz noted that we communicate with the outside world through movement trajectories. Natural movement can be decoded as a velocity, in which the trajectory is a behavior output. For example, researchers can derive a vector based on the firing of a group of 300 neurons and decode the intended movement direction. These vectors can then be used to control robotic arms and other devices. The group has achieved this goal in monkeys: with an electrode array implanted in its brain, a monkey can use a robotic arm to grab food and place the food in its mouth. This technology could allow paralyzed people to regain control of their limbs.

Developing algorithms for enhanced brain–machine interfaces

"It turns out that the population vector algorithm is a terrible statistical estimator," began Steven M. Chase from Carnegie Mellon University in his talk on the algorithms researchers now use for vector analysis. The population vector algorithm is biased and does not accurately assess the intended movement direction. In 1994, Salinas and Abott described another algorithm that would compensate for this bias, the Optimal Linear Estimator (OLE).

The OLE takes input from the same neurons; however, the neurons push in slightly different directions from their own preferred direction, which helps to compensate for non-uniformity in the population. In an off-line simulation OLE outperforms PVA, but OLE and PVA perform similarly on-line because subjects can compensate for bias, at least to some degree and in some circumstances. Therefore, restricting development of new technologies to the class of unbiased estimators may not be necessary for accuracy in prostheses. However, it remains to be determined whether there are unbiased decoders that might outperform their biased counterparts on-line. Chase concluded by offering advice for algorithm development in new technologies: It is not necessary to restrict development to the class of unbiased algorithms. It is possible to multiplex multiple dimensions of control from a high-fidelity source. Off-line assessments of which algorithms work better may not accurately predict on-line performance. To make real progress in algorithm design we have to understand the learning process.

Bias in the population vector algorithm. (Image courtesy of Steven M. Chase)

Neurotechnology to predict, prevent, and control seizures

Brian Litt from the University of Pennsylvania has studied similar technology in epilepsy. Epilepsy affects 60 million people globally and over one-third of patients are treatment-resistant. Litt's team found that the brain activity preceding a seizure—previously thought to be decrement, decreased activity—is actually high-frequency activity. It records as limited-frequency activity because our technology is inadequate. They implanted an 8 cm × 8 cm series of electrodes, with 1 electrode impacting 12 million neurons. With stimulation, seizures decreased by 40% compared to sham controls.

The goal of this work is seizure freedom. The treatment is not yet approved by the FDA but it is approved in the EU. Litt likens the current electrode implants to the first pacemakers but acknowledges that our understanding of the neural networks involved in seizures is far from complete. The technology provides evidence of some efficacy. Future challenges will be to better define seizures and to find new means of suppression.

High-density neural sensors that can detect the two-dimensional structure of a seizure may allow us to detect signal warnings and could be used to realize seizure freedom. Multiscale sensors and effectors are also needed, and sampling error and scale problems remain to be solved. As new tools such as organic electrodes and nanotechnology become more sophisticated this technology will improve, but it will probably always be impossible to sample the entire brain so we must determine the appropriate scale at which to work. Litt believes that we need to combine neuroscience, computational modeling, hardware, data mining, and algorithms in an academic, clinical, and industry collaborative approach. Achieving these goals depends on data sharing, agreed-upon data standards, and a team-based approach to science.

Chair:
David Eidelberg, Feinstein Institute for Medical Research
 
Speakers:
Susan M. De Santi, GE Healthcare
Arthur Toga, University of California, Los Angeles

Highlights

• Neuroimaging advances may allow for biomarker discovery and development, which will help researchers test new treatments for diseases like Alzheimer's disease.
• There are several new imaging technologies in the testing and validation stage.
• With advances in imaging, there is a need to aggregate and analyze large quantities of newly available data. This will require a collaborative effort among large working groups of researchers.

Imaging applications in translational neuroscience

David Eidelberg from the Feinstein Institute for Medical Research introduced advances in neuroimaging. A variety of imaging tools have been developed to evaluate anatomical, functional, and neurochemical changes associated with brain disorders. These tools often have signal-to-noise (SNR) characteristics that are sufficient for reliable assessments in single subjects. However, we also need computational algorithms that can assess systems-level (network) changes in progressive neurodegenerative disorders.

Researchers are working to discover and implement network biomarkers that will improve diagnostic accuracy and assess disease progression and treatment responses in individual subjects. Cognitive and behavioral changes in brain disorders are mediated by a variety of mechanisms involving multiple neurotransmitters; for example, dementia in Alzheimer's disease is not simply associated with cortical amyloid deposition. It is unknown whether or how these systems can be manipulated therapeutically. It is also unknown whether such biomarker measures are sufficiently sensitive to be used to follow progression in individual subjects or whether functional network biomarkers will have sufficient sensitivity and specificity for the classification of individual cases. Moreover, diagnostic covariance patterns may or may not be sensitive to disease progression and treatment responses. Thus, the optimal imaging modality or analytical algorithm for systems-level assessments is not known. Much will depend on further developments in magnetic resonance imaging (MRI) technology, positron emission tomography (PET) technology, and optimization algorithms.

Biomarkers to predict Alzheimer's disease. (Image courtesy of David Eidelberg)

Improving diagnostic accuracy and the sensitivity of potential biomarkers of disease progression requires an emphasis on imaging modalities demonstrating high replicability across populations, scanning sites, and imaging platforms. Algorithms to enhance the accuracy of individual subject measurements need to be optimized. Emerging technologies and analytical approaches need to be evaluated for feasibility and value-added in a multicenter trial environment. Imaging metrics incorporated as potential outcome variables in clinical trials will provide critical information about the natural history of disease and the effects of treatment and placebo, both at the regional level and at the network level. This approach is likely to be especially valuable in gauging progression in individuals with preclinical disease.

Imaging biomarkers for prediction and progression of neurodegenerative disease

Susan M. De Santi from GE Healthcare presented imaging work in neurology at GE. The company has FDA-approval for ioflupane I-123 injection (DaTscan), a radiopharmaceutical indicated for striatal dopamine transporter (DaT) visualization, in single photon emission computed tomography (SPECT) brain imaging to assist in the evaluation of adult patients with suspected Parkinsonian syndromes (PS) and essential tremor. The compound can be used to help differentiate essential tremor from tremor caused by PS.

Other PET tracers include the amyloid imaging agent flutemetamol F18 injection (now FDA approved, Vizamyl); the neuroinflammation marker under development, GE-180; and the tau protein marker in development, GE-216. [¹⁸F]Flutemetamol has a short half-life (~110 min) and must be manufactured, delivered, and used within in a single day from NDA-approved manufacturing sites. Production is facilitated by the use of an on-site cyclotron and FASTLab™ PET synthesizer. For approval of amyloid imaging agents, regulatory agencies require autopsy studies to demonstrate the association between imaging and pathological changes in the brain. The flutemetamol phase III pivotal trial recruited end-of-life patients and compared flutemetamol PET imaging results to postmortem autopsy in subjects who died within one year of imaging. Flutemetamol F18 imaging correlated well with fibrillar amyloid pathology based on a modified Bielschowsky silver staining score.

Imager training is required to read the color flutemetamol images. In a trial of trained readers, intra-reader agreement was 91%, with high rates of sensitivity and specificity. GE found some false negatives in subjects with brain atrophy and some false positives in subjects with a high concentration of diffuse plaque, which is not detected via Bielschowsky staining. GE is conducting a clinical trial in patients in the prodromal stage known as amnestic mild cognitive impairment (aMCI). The ongoing 3-year study aims to determine whether a positive flutemetamol scan predicts progression to AD. In a separate, small phase II trial, 7 of 9 aMCI patients with a positive flutemetamol scan progressed to AD over 2 years compared to 1 of 10 with a negative flutemetamol scan.

New neuroimaging tools for understanding and predicting neurological disease

Arthur Toga from the University of California, Los Angeles, discussed the need to find AD biomarkers at early time points, so that disease progression can be slowed. AD is at the forefront of research because of the rapidly aging population, but strategies to find biomarkers for AD may also be useful for other neurodegenerative diseases. "[Researchers should not] take one single measurement as a panacea. All kinds of data need to be combined together to create a signature either of a kind of disease or some phase of an individual" Toga said.

His group looks at aggregate data to find patterns that could predict disease progression. They can create probabilistic representations (i.e., brain atlases) from large data sets for potential use in clinical settings. One of the challenges is to combine data from different sources. All brains are different, but the atlases need to be able to go from the individual to the population level without a loss of detail. Ultimately, the group aims to create a wiring diagram of the brain that correlates with function. The wiring diagram would be sensitive to variations in disease cohorts and normal individuals, as well as variations in normal development and aging.

To create a robust brain atlas, scientists need to crowdsource: high volumes of data are required. One arm of this project is the Alzheimer's Disease Neuroimaging Initiative (ADNI), which validates the use of biomarkers including blood tests, cerebrospinal fluid tests, and MRI/PET imaging for AD clinical trials and diagnosis. There are 58 acquisition sites for ADNI. The aim is to create a connectome and to discern differences between normal neural connections and "connectopathies." Toga's group has 4 petabytes of storage and 4000 processors running in parallel to aggregate and integrate the massive amount of data they are collecting.

Chair:
Helen Mayberg, Emory University
 
Speakers:
Jerrold Vitek, University of Minnesota
Donald Malone Jr., Cleveland Clinic
Edward Boyden, Massachusetts Institute of Technology

Highlights

• Deep brain stimulation has been successful in mitigating depression in some patients with movement disorders, seizures, and neuropsychiatric disorders, but both the technology and the targets need to be optimized.
• New technologies like optogenetics—the introduction of light-sensitive genes into neurons, allowing for light-based control of neuronal pathways—hold significant promise in neuromodulation for disease therapy.
• Optogenetics is still in its infancy and much work needs to be done before it can become a common tool for the treatment of neurological diseases.

Focal modulation of disease circuits: where and how, now and future

In her introduction to the session, Helen Mayberg of Emory University explained that advances in neuromodulation require bidirectional translational work. Progress in deep brain stimulation (DBS) was catalyzed by advances in stereotaxic neurosurgery and implantable stimulation delivery systems, the availability of structural/functional imaging and electrophysiology tools, and the development of circuit models of disease and rational targets for focal stimulation. DBS is currently used to treat movement disorders, seizures, and neuropsychiatric disorders. Although circuit-level modulation for human disease is available, the technology—while clinically effective—is relatively crude.

The next generation of existing devices will require better flexibility and control and more precise unit recordings and anatomical targeting. A shift in focus to network dynamics, scaling up from current targets, will hopefully lead to smart stimulation systems that use high-resolution, multifunctional real-time readouts. Goals such as cell-specific targeting require "reverse translation," moving research from the bedside to the bench. Because of behavioral and anatomical differences between animals and humans, animal model systems may not be adequate surrogates. As with most neurotechnology trials, safety, regulation, implementation, logistics, and ethics must be considered.

Optimization of deep brain stimulation. (Image courtesy of Helen Mayberg)

Deep brain stimulation for Parkinson's disease: challenges and opportunities

Jerrold Vitek of the University of Minnesota discussed the use of DBS for Parkinson's disease (PD). Current challenges include defining a pathophysiologic target, localizing the device (position imaging and refining location through microelectrode recording), and modulating disease activity. Vitek agreed with Mayberg that these challenges need to be met through a combination of basic science and clinical trials.

A better understanding of both pathophysiology and DBS has expanded our targets for PD, with animal models providing the rationale for target selection. Advances in lead technology, including directional split-band electrodes and leads that enable spatial-temporal patterning, will allow for a "sculpted" current, in which electric-current fields can be distributed to optimize impact. Microelectrode recording could improve electrode placement. Optimizing current fields may improve treatment efficacy, battery life, plasticity induction, production of trophic factors, and duration of effect. Future DBS technology will ideally involve very small cranial neurostimulators placed on top of the head, with no tunneling or extensions required. These devices would allow for wireless telemetry, could be recharged without a vest or other device, and could be controlled through a closed-loop system. Vitek predicted that these devices could begin to appear within 5 years.

Understanding pathophysiology in Parkinson's disease improves target selection for deep brain stimulation. (Image courtesy of Jerrold Vitek)

Recent developments in clinical trials using deep brain stimulation for depression

Donald Malone Jr. of Cleveland Clinic reviewed recent developments in DBS clinical trials for treatment-resistant depression (TRD). The best patients for DBS are selected based on an accurate diagnosis, sufficient severity of illness, nonresponse to less-invasive options, and ability to provide informed consent. DBS has been used in obsessive-compulsive disorder (OCD), based on lesion studies and knowledge of neurocircuitry. The initial findings for OCD by Nuttin and colleagues showed a 50% improvement, leading to FDA approval under humanitarian device exemption (HDE). DBS is indicated for chronic, severe treatment-resistant OCD in adults who fail at least 3 adequate SSRI trials.

Various DBS targets have been tried for TRD, including ventral capsule/ventral striatum (VC/VS), which is based on OCD trials and functional imaging studies. Along with Cg25, these are the most studied target regions in controlled clinical trials. Other open-label studies have investigated the nucleus accumbens, inferior thalamic peduncle, lateral habenula, and medial forebrain bundle. In 2011, Kennedy and colleagues reported on Cg25 DBS in 20 patients who were followed for 3–6 years. The responses to treatment at years 1, 2, and 3 were 62.5%, 46.2%, and 75.0%, respectively. The response at the last follow-up (LFU) was 64.3%. Holzheimer and colleagues studied Cg25 DBS in 10 major depressive disorder (MDD) and 7 bipolar affective disorder (BAD) patients with sham lead-in. Subjects had response rates of 41% at 24 weeks (N=17), 36% at 1 year (N=14), and 92% at 2 years (N=12). The ongoing BROADEN study is evaluating Cg25 as a target in a double-blind, sham-controlled trial.

Malone and colleagues conducted a VC/VS pilot study from 2003 to 2007 in 17 TRD patients. At LFU, 71% were responders and 35% remitted, with a follow-up range of 14–67 months. Two additional recent trials have examined a VC/VS target for DBS. The RECLAIM trial followed TRD subjects and found no difference between sham subjects and those with VC/VS DBS targeting. However, both groups demonstrated decline on the Montgomery–Åsberg Depression Rating Scale (MADRS). As might be expected in a study of TRD, most adverse events were associated with depressive disorder. The lack of difference from sham in the RECLAIM study could be explained by several factors, including true lack of efficacy, a lesional effect with sham surgery, protection of the blind, inadequate blinded period, and patient selection. This study calls into question whether the right outcomes are being measured and whether measures to determine a treatment response are set too high.

Optogenetics and other tools for analyzing and controlling neural circuits

Edward Boyden of Massachusetts Institute of Technology discussed another method to modulate neurons: optogenetics, the process of using optics to modulate neurons by converting light to electricity. This is accomplished through the introduction of genes for nonspecific cation channels that open in response to light. Using gene therapy to insert these genes into the plasma membranes of target cells could allow for the production of action potentials in response to light stimulation. The channel proteins are opsins, seven-pass transmembrane proteins discovered in bacteria. Three known classes of these proteins are light-sensitive: channelrhodopsins, halophodopsins, and archerhodopsins. By targeting specific cell types, optogenetics can be used to modulate the excitation of neighboring cells. This tool may also be useful for photoreceptor degenerative diseases, such as retinitis pigmentosa. Individual light-sensitive opsins respond to different wavelengths, so current work focuses on using multiple color optogenetics to allow for more refined modulation of cell types. Arrays of hyper-dense optical fibers could enable greater control of neuronal cell types. There is still a need to determine how best to regulate neural circuits.

Chair:
Ole Isacson, McLean Hospital; Harvard Medical School
 
Speakers:
Arnaud Lacoste, Novartis Institutes for BioMedical Research
Ivar M. Mendez, University of Saskatchewan and Royal University Hospital, Canada

Highlights

• The development of induced pluripotent stem cells has allowed for new approaches to researching the mechanisms of action for Parkinson's disease and its treatment.
• Stem cell transplants for Parkinson's disease patients have resulted in significant reduction of long-term morbidity for some patients in clinical trials.

New stem cell technologies considered for applications to brain diseases

Ole Isacson of McLean Hospital and Harvard Medical School described the use of stem cells in Parkinson's disease (PD). Isacson noted that the onset of symptoms of PD does not mark the beginning of the disease. Instead, PD is a threshold disease, where the underlying pathology is finally revealed in symptoms once a minimum loss of neurons is reached. PD results from the loss of neurons in the substantia nigra pars compacta (SNC or A9) region of the brain. Before the minimum-loss threshold is reached, plasticity and a compensatory network in the brain mask the disease. Thus, there is an effort to intervene in PD before symptomatic disease presents, but biomarkers that predict disease are needed to accomplish this goal. The lack of a biomarker for early disease may have contributed to the lack of success in neuroprotective trials for PD.

Our understanding of PD has been improved through the use of induced pluripotent stem cells (iPSCs) that have been reprogramed/deprogramed to become totipotent. Totipotent cells can function as stem cells. When taken from PD patients, iPSCs can be used to create a "disease in a dish" model of PD, with iPS cells programmed to become dopaminergic neurons, a key type of neurons lost during disease progression. Isacson and colleagues used this model to study mitochondrial function in PD. They found that cells with a mutation in PINK1, a PD-related protein, had increased production of free radicals following cell depolarization. The researchers could rescue mitochondrial function in these cells with pharmacologic agents.

Reprogrammed cells also have the potential to replace lost cells in the brain, and there may be a way to restore circuitry. This technique is advantageous because it uses the patient's own cells, helping to circumvent the immune response that would occur using non-autologous cells.

Arnaud Lacoste from Novartis Institutes for BioMedical Research discussed how iPSC studies are being used in research and development to identify targets and potential therapeutics. He explained how industry is using iPS cells to find targets for retinitis pigmentosa.

Clinical cell transplantation for Parkinson's disease: surgical techniques and methodology

Imaging differences following fetal stem cell implantation for PD patients. (Image courtesy of Ivar M. Mendez)

Ivar M. Mendez from the University of Saskatchewan and Royal University Hospital expanded on the session topic, focusing on clinical fetal stem cell transplantation in patients with PD. This technique seeks to create connections through reinnervation. Animal models can be useful for optimizing cell-replacement methods in studies of cell survival and reinnervation, surgical delivery, immunosuppression, target regions, imaging, and clinical assessment. If successful, a new technique is moved into small open-label trials for dopaminergic grafts in patients.

Mendez outlined the procedure for surgical cell grafts, which includes imaging to develop an operating-room plan for the patient graft and a clinical injector to transplant the cells. Post-operative imaging after 4 or more years showed an increase in activity compared to pre-operative activity. Mendez shared a video of a patient 8 years after a bilateral transplant—the patient has been off medication and is performing as well as the best DBS PD patients. Fetal stem cell transplantation has advantages over DBS: the substantia nigra continues to degrade in DBS but there is reinnervation with cell transplantation. However, the cells must be positioned in the correct location.

Keynote Speaker:
Robert J. Greenberg, Alfred E. Mann Foundation
 
Moderator:
Orla M. Smith, Science Translational Medicine
 
Speakers:
Michel Goldman, Innovative Medicines Initiative
Ron Maron, Israel-U.S. Binational Industrial Research and Development (BIRD) Foundation
Justin C. Sanchez, Defense Advanced Research Protects Agency (DARPA)

Highlights

• For new neurotechnologies to reach the market and have a clinical impact, several development barriers must be addressed, including cost and regulatory approval.
• Some barriers can be limited through collaboration between small companies, which often develop the initial technologies, and larger entities that can bring them to market.
• Collaboration between researchers, industry, and other stakeholders will quicken the development cycle and improve the likelihood that a new technology is commercially available.

Keynote address: a retinal prosthesis — from idea to clinical application

Robert Greenberg from the Alfred E. Mann Foundation delivered the keynote address, on the development of a retinal prosthesis. In the early 1990s, Greenberg and colleagues at Johns Hopkins University found that they could electrically stimulate the retina in blind patients with retinitis pigmentosa (RPE), producing the perception of light. Photoreceptors are degraded in RPE patients, but the underlying retinal bipolar and ganglion cells survive in large numbers. A multidisciplinary team has now developed an implantable retinal prosthesis that stimulates these cells. The resulting device, the Argus II Retinal Prosthesis System, obtained approval in Europe in 2011 and U.S. FDA approval 2 years later. Obstacles to overcome in the development process included generating capital, optimizing the product, gaining regulatory approval, and developing manufacturing procedures. The device costs $163 000 in the E.U. and $100 000 in the U.S.

The prosthesis has a thin-film electrode array that is implanted into the eye. The array is coupled with a pair of glasses that contain a miniature video camera, a transmitter coil, and a small wearable computer that translates the image to the electrode array. This allows for the production of an image on the retina. Although the image only contains a limited amount of information compared to full sight, the system restores some functionality to completely blind patients. Users can find doors, follow sidewalks, and identify the direction of a person walking in front of them, and they have about a 20° field of vision. These functions can significantly improve users' daily lives. Fifty-six patients who previously had no treatment options have received the implant. Greenberg noted that improvements need to be made to generate real images; the best option to do so is to increase image resolution through better software.

Keynote panel: accelerating neurotechnology research — a global perspective

Orla M. Smith from Science Translational Medicine moderated the keynote panel, which focused on how to accelerate neurotechnology research worldwide. Several large neuroscience collaborations are underway around the world, including the Human Brain Project in the EU and the BRAIN Initiative in the U.S. The panelists discussed their experiences fostering collaboration between a diverse set of organizations interested in neurotechnology. Michel Goldman has been involved in 47 projects for the Innovative Medicines Initiative, a European organization designed to stimulate multi-stakeholder collaboration. He pointed to the scope of current projects and the amount of data generated as reasons for neuroscience to function collaboratively but acknowledged that developing trust between organizations is challenging.

Ron Maron also works to build collaboration as part of the Israel-U.S. Binational Industrial Research and Development (BIRD) Foundation. The foundation stimulates U.S.–Israeli projects with grants, which are repaid if revenue is made. The foundation has supported approximately 850 projects in diverse fields, including cybersecurity, electronics, and life sciences. "Collaboration is an art," he said; it is necessary to manage, monitor, and guide collaborations to produce success. Smaller organizations tend to move more quickly than their larger counterparts, and this can be a barrier to effective collaboration.

Justin C. Sanchez has an international focus on neuroscience and neurodevelopment in his role at the Defense Advance Research Projects Agency (DARPA). He said that there is, in general, a very fragmented approach to the brain and argued for better coordination to advance the next generation of research. A "unified pull/push" between industry and academia can help to deliver results with a high potential for payoff. As collaborations grow, interaction between various initiatives should happen organically. However, collaboration will be hampered as long as academic promotion criteria are tied to individual output. Maron agreed that the hierarchy of professors makes it more difficult to collaborate with academics. The panelists also noted that joint projects cannot be the sole means to advance neuroscience: individual discoveries also drive progress.

Speakers:
David M. Holtzman, Washington University School of Medicine
Sean Mackey, Stanford University
Helen Mayberg, Emory University
David Eidelberg, Feinstein Institute for Medical Research
Ann C. McKee, Boston University School of Medicine

Highlights

• Across disease states, there is a need for improved methods of diagnosis and for biomarkers that can predicts disease progression. Advances will be limited until validated biomarkers are identified and developed.
• Advances in technology longevity and functionality are much sought after in multiple disease treatments.

Introduction

A series of short state-of-the-disease lectures set up breakout sessions on specific topics in neuroscience and neurotechnology. The lectures covered PD, depression and mood disorders, chronic pain, AD, and traumatic brain injury. There was also a breakout session for neuromuscular control/neuroprosthetics. Breakout groups were asked to consider the following questions:

• What specific gaps in knowledge and elsewhere in the translation process are barriers to the diagnosis and treatment of your disease/disorder of interest?
• What next steps are necessary to fill these gaps and accelerate the development of technology or other tools for improved diagnosis and treatment?
• What objective metric(s) could be used to gauge the success of these technologies and treatments (e.g., biomarkers)?

Alzheimer's disease: what is known and where to go

David M. Holtzman of the Washington University School of Medicine introduced the topic of Alzheimer's disease. As the U.S. population continues to age, there is a looming AD crisis. The expected cost of AD in 2013 is $200 billion. AD is a disorder of protein misfolding, in which amyloid-beta aggregates and insoluble tau protein increase inflammation and toxicity in the brain. The build-up of these proteins correlates with disease progression, but there is still a need for markers that predict disease prior to the development of symptoms. Cerebral binding of the amyloid-binding molecule, [11C]PIB, is a candidate biomarker of AD that could be detected through imaging. Additionally, CSF tau and Aβ42 predict progression from normal to very mild dementia/mild cognitive impairment (MCI), as well as from MCI to AD. There are many treatments in development, but in the absence of good predictive biomarkers for disease progression it will be hard to effectively evaluate treatments for presymptomatic AD. Whole-genome profiles, small-vessel imaging, and knowledge of environmental influences on AD would all be helpful in further defining presymptomatic disease and the risk of progression.

The current model of Alzheimer's disease brain pathology (Image courtesy of David M. Holtzman)

Chronic pain: brain imaging biomarkers for pain

Sean Mackey of Stanford University introduced the topic of chronic pain. In the U.S., pain is the top reason that people are unable to work and Vicoden (hydrocodone/acetaminophen) is the most commonly used medication. Chronic pain is affected by anxiety, mood, catastrophizing, somatization, early life experiences, and other factors. Self-reported pain remains the gold standard for pain assessment, indicating a need for objective biomarkers. There are patient populations who cannot accurately indicate their level of pain, and brain imaging may be able to help these patients. Neurological imaging signatures of pain are under development, but improved targets are needed. Further, current preclinical models are not always applicable to humans and treatment targets require longitudinal assessment. It is better to prevent pain than to treat it, and objective biomarkers will help in the development of pain-prevention techniques. Diagnosis faces challenges because there is a large overlap between chronic pain symptoms and other disease states. Further understanding of comorbid conditions is needed to advance pain research, as is improved understanding of the mechanisms of chronic pain. We need to identify objective markers and patient phenotypes, perhaps even for different subsets of patients who may experience pain differently, such as children and adults.

Activated areas of the brain during nociceptive pain. (Image courtesy of Sean Mackey)

Major depression and mood disorders

As noted by Helen Mayberg of Emory University, treatments that are available for depression and mood disorders do not work for all patients. How, then, can doctors find the right treatment for the right patient? There is a need to limit relapse, to develop new interventions and targets, and to identify how current treatments converge. Defining subsets of patients for particular treatments may require additional biomarkers, as well as information on the circuits involved in these disorders. Disease heterogeneity is a primary barrier to treatment, but improved diagnostic criteria could alleviate problems caused by patient variability. Heterogeneity in disease also makes it difficult to develop appropriate animal models. In clinical trials, we need to look at multiple timepoints to account for increased variability caused by environmental factors.

DBS goals for treatment-resistant depression. (Image courtesy of Helen Mayberg)

Neuromuscular control and neuroprosthesis

There are significant gaps neuroprosthetic technology. Device longevity is a particularly big concern because degradation over time reduces efficacy and necessitates replacement. Work to increase device longevity is part of an effort to understand the needs of patients, which must be considered during device development. Better collaboration between clinicians and engineers will improve development, and better collaboration between developers, regulators, and payors will ensure that products can succeed in the marketplace once approved.

Parkinson's disease

David Eidelberg of the Feinstein Institute for Medical Research presented on PD. Several processes contribute to PD, but they appear to converge at the organelle level. Much current work focuses on determining networks and targets for stimulation. Better correlation between networks and clinical status is needed, requiring the development of objective imaging measures to define the prodromal phase. Current knowledge gaps in PD are enormous. It is known that there is a decline in dopamine over time until the disease becomes symptomatic, but the prodromal state must be characterized and therapies that impact outcomes are needed. However, patient variability makes this work particularly challenging. Better metrics and biomarker standardization are the most pressing needs for drug development.

Network analysis in Parkinson's disease. (Image courtesy of David Eidelberg)

Traumatic brain injury: emerging concepts in chronic traumatic encephalopathy

Ann C. McKee of Boston University School of Medicine discussed chronic traumatic encephalopathy (CTE). CTE has two clinical presentations: behavioral/mood disorders and cognitive impairment. There is a unique pathology to CTE, including brain atrophy and tau protein deposits. CTE is increasingly categorized as an acquired frontotemporal lobar degeneration and progresses slowly. The disease is not currently diagnosable but can be identified postmortem. Objective measures of concussion are needed for diagnosis and to determine incidence and prevalence. Diagnosis of incident CTE is the "holy grail" of research on the disease. Ideally, diagnosis would be possible through a blood test, but regardless of its form, the test needs to be objective and validated. Neurofilament light chain may be a candidate biomarker. Once a diagnostic tool is developed, there will need to be better treatments and return-to-play guidelines. Clinicians and researchers also need to recognize that CTE is a chronic injury of neuroinflammation.

CTE progression. (Image courtesy of Ann C. McKee)

Speaker:
Cristin Welle, Center for Devices and Radiological Health, FDA

Panelist:
Kristen Bowsher, Center for Devices and Radiological Health, FDA

Highlights

• Regulatory challenges remain a hurdle for any new device, but understanding the FDA process prior to development can lessen the time to market.
• The FDA seeks to be transparent and to partner with developers to help them navigate the regulatory process as easily as possible.

Regulatory hurdles for the development and use of medical devices

Cristen Welle from the Center for Devices and Radiological Health (CDRH), FDA, provided an overview of how regulatory process within the FDA apply to neurotechnology. Evaluation for these devices falls mainly under the CDRH, specifically the Division of Neurological and Physical Medicine Devices. There are three paths for a medical device to come to market, depending on the class of the device. Class I devices are low risk and are exempt from most of the FDA requirements that class II and class III devices must meet. Class I devices include such items as manual surgical instruments, tuning forks, neurosurgical chairs, external limb prosthetics, and orthotics. Some class I devices must submit a 501(K) for clearance, to evaluate whether the device is substantially equivalent to a predicate.

Class II devices, which include items such as EEG, cutaneous electrodes, shunts, clot retrievers, and motorized wheelchairs, must also show that they are equivalent to a predicate. Class III devices, such as stents and devices for vagus nerve stimulation, spinal cord stimulation, and deep brain stimulation must go through pre-market approval, and developers should determine which submissions should be filed and whether clinical data are required. Developers can send pre-submission inquiries and receive feedback from the FDA. In some cases these inquiries require an investigational device exemption (IDE), allowing the investigational device to be used in a clinical study in order to collect data on its safety and effectiveness.

FDA pathways for development of medical devices (Image courtesy of Cristen Welle)

Welle and Kristen Bowsher, also from the CDRH, ended the session with a panel discussion. Welle noted that the FDA increasingly conducts its own basic neuroscience programs, focusing on neural implants and functional performance. Research is geared toward creating standards for new technologies and the investigators are encouraged to publish their results and are allowed to collaborate with external scientists.

The FDA has actively worked to reduce antagonism with developers by improving transparency and collaboration. They have also developed new programs to this end, such as Innovation Pathway and Entrepreneurs-in-Residence.

Presenters:
Xinyan Tracy Cui, University of Pittsburgh
Dennis A. Turner, Duke University
James Phillips, University of Washington

Highlights

• Academic researchers contribute significant advances to the field, particularly toward our understanding of the strengths and limitations of neurotechnologies that are in use.
• By partnering with industry and leveraging existing technologies, researchers can rapidly develop and bring new technologies to market.

Bio-mimetic neurotechnologies camouflage implants as neurons: pioneering new challenges in FDA translation

In the submitted abstracts session, Xinyan Tracy Cui from the University of Pittsburgh presented on biological responses to implants. When a device is implanted in the brain, microglia, other glial cells, and macrophages encapsulate it. This inflammatory tissue response results in a decrease in neural density, thereby limiting the device’s function. As a result, electrode recording will degrade over time. To surmount these challenges, Cui coated electrodes with the neuronal-specific adhesion molecule L1, which inhibits glial cell attachment and allows for axonal growth towards the probe, improving the function of the implanted device over time.

Thalamic microstimulation for somatic sensory substitution

Dennis Turner of Duke University presented on thalamic microstimulation for somatic sensory substitution. Sensory substitution seeks to replace lost sensation through a restorative interface that inserts external sensory information directly into the CNS. Turner and colleagues investigated microstimulation of the thalamus; the sensory thalamus is dense, so stimulation is safe and effective via a frontal entry point, and stereotactic planning can be used. Using microwire arrays, they produced subliminal responses in small areas such as the finger. Many of the responses were "natural"; that is, they resembled pressure or touch, or tingling, and none were reported as painful. Increasing the intensity did not lead to stronger stimulus, but did increase the area of response. Comparatively, DBS with macroelectrodes produced a larger area of stimulation (for example, 2 to 3 fingers). Different contact points gave only slightly different areas of stimulation, with less spread than wires. The responses were less "natural" than microwires, while patterning and flutter minimally altered sensation.

Electrodes for thalamic stimulation (Image courtesy of Dennis Turner)

Vestibular prosthesis: five years from design, to monkey, to man

James Phillips from the University of Washington reported his experience bringing a vestibular prosthesis for the inner ear from design to human use in 5 years. The device is intended to correct vertigo in patients with bilateral loss of vestibular function, uncompensated unilateral loss of vestibular function, or fluctuating balance function (such as in Meniere's disease). Building on cochlear implant technology, his team created new intellectual property by modifying the hardware and software to create appropriate stimulation. Partnering with an existing cochlear implant manufacturer helped the development process significantly, providing his team with an experienced party to negotiate the clinical development process and avoid barriers at the FDA. The device was evaluated in 11 rhesus macaques and then moved to clinical trials in Meniere's disease patients. The team submitted a pre-IDE to the FDA in collaboration with their industry partner, met with the FDA to review, and then submitted the full IDE. The device produces slow-phase eye movements, body sway, and perceived motion, which is consistent with normal vestibular activation. Philips noted that the rapid development of the device was aided by industry collaboration, the ability to leverage existing technology, and the availability of an appropriate animal model to test production devices.

What are the best ways to stimulate collaboration between stakeholders in neurotechnology advancement?

How can collaborations be enacted and the interests of the partners be aligned?

Are there undiscovered biomarkers that could be used for the diagnosis and prediction of neurological diseases?

How can we best identify and validate novel biomarkers?

How can we improve the interface between technology and the body?

How can we improve device longevity and functionality?

Are ethical considerations being addressed alongside neurotechnology advancement?

What are the best targets for deep brain stimulation?

How can we manage the massive amounts of imaging data being generated?

What are the most appropriate algorithms to use when decoding neuronal networks?

Are there less invasive ways to modulate neuronal control?

What are the best ways to spur neurotechnology advances?