Nuerorobotics: Development of Brain-Machine Interface Devices

1. CPG Hip Knee Ankle NEUROENGINEERING RESEARCH P R O G R A M O V E R V I E W Nuerorobotics: Development of Brain-Machine Interface Devices • The project is directed toward understanding how sensory information is processed by the brain's neural system and is then transformed into an appropriate motor output. The new microelectrodes are designed on ceramic wafers and might be able to increase the yield of recordings by an order of magnitude for long periods of time (ultimately years). Biomimetic Legged Robot Locomotion • The goals of this research project are to design a neuron-like central pattern generator (CPG) for control of legged locomotion and to implement this CPG into legged robots. Neural Control in the Spinal Cord & Motor Cortex: Modularity and Force-field Primitives • This project focuses on modularity of control in biological motor systems. The goal is to describe neural circuits and associated motor actions organized in the spinal cord and motor cortex as collections of visco-elastic force-field primitives. These visco-elastic force-field primitives form a basis to construct both reflexes and novel, learned motions and interactions. Intraspinal Microstimulation for Motor Neural Prostheses • The interneuronal circuitry of the spinal cord is capable of generating complex movements with coordinated muscle activity. Current motor neuroprostheses are based on stimulation of the last order neurons, thereby bypassing this organization. A neural prosthesis based on the activation of spinal circuits may improve the control of complex multi-joint movements, thus providing an increase in functions to individuals with spinal cord injury. Informatics Center for Mouse Neurogenetics • The purpose of this project is to develop and exploit a suite of image databases, motorized Internet microscopes, and software to study the genetic basis of structural variation of the mouse CNS. The focus of this project is on genes responsible for the extraordinary variation in CNS structure among mice and humans. Computational Modeling of Neural Control of Breathing • The major objective of this project is to build a comprehensive computational model of neural control of breathing that is consistent with experimental data from the cellular to the systems levels. The neural control of breathing is considered as a model for computational study of cross-level integration of cellular, network, and systems neural mechanisms in the brain's motor control. Computational Modeling of the Spinal Cord Neural Circuitry (Neural Control of Locomotion) • The ultimate goal of this research is to build a comprehensive model of spinal cord neural circuitry which integrates basic reflexes and neuronal central pattern generators for limb control during locomotion. • Faculty: Ilya A. Rybak, Ph.D., Research Professor in BIOMED, Drexel University; Karen A. Moxon, Ph.D., Assistant Professor in BIOMED, Drexel University. • Collaborating Researchers/Graduate Students:Refer to individual Project 1-PAGERS that follow this Program Overview.

2. The new microelectrodes are designed on ceramic wafers and hold the possibility of being able to increase the yield of recordings by an order of magnitude for long periods of time (ultimately years). Ceramic-Based Multi-Site Electrodes (CBMSE) Patent Pending NEUROROBOTICS: DEVELOPMENT OF BRAIN-MACHINE INTERFACE DEVICES P R O J E C T O N E P A G E R • The project is directed toward understanding how sensory information is processed by the brain's neural system and transformed into appropriate motor output. • Methods: • Computational modeling • Neural ensemble recordings using recently developed • novel methods for interfacing with the brain. • Critical problems: • The neural recordings are stable for only a short time period. • The electrodes are too large to record from single neurons. Dr. Moxon received her MS and Ph.D. from the University of Colorado, Dept. of Aerospace Engineering. She has worked with leading neuroscientists over the past several years to investigate how ensembles of neurons in the brain represent information and to use that information for the development of novel brain-machine interface devices, such as neuroprosthetic devices, a neural control device for epileptic seizures, tremor control in Parkinson’s disease, and seizure detection in epilepsy patients. • Faculty/Contact: Karen A. Moxon, Ph.D., Assistant Professor in BIOMED, Drexel University; Adjunct Assistant Professor, MCP Hahnemann School of Medicine, Dept. of Neurobiology & Anatomy. • E-mail: Karen.Moxon@drexel.edu • Collaborators: John K. Chapin (SUNY, Brooklyn) and Simon Giszter (MCPHU). • Support:NIH and DARPA.

3. Research Objectives: • Design a neuron-like central pattern generator (CPG) for control of bipedal locomotion. • Implement the CPG for locomotion control of       bipedal humanoid robot (Iguana Robotics,       Inc.). The creation of bipedal walking robots is one of the most challenging tasks in modern robotics. The proposed work is a necessary step toward the creation of a new generation of such robots. Hip Knee Ankle Approach: Our approach is based on the explicit implementation of the model of neural control of locomotion in mammals for real-time control of bipedal robots. The biomimetic control architecture, developed and implemented in real robot-bipeds, will use proprioceptive feedback and sensory (including visual and vestibular) information for automatic adaptation of locomotor behavior to physical characteristics of the legs (as well as the rest of the entire body), characteristics of the environment, and behavioral goals. Visually guided robot-biped, Iguana Robotics, Inc. BIOMIMETIC LEGGED ROBOT LOCOMOTION P R O J E C T O N E P A G E R • Faculty/Contact: Ilya A. Rybak, Ph.D., Research Professor in BIOMED, Drexel University. • E-mail: rybak@cbis.ece.drexel.edu • Collaborators: M. Antony Lewis (Iguana Robotics, Inc.); John K. Chapin (SUNY, Brooklyn); Boris I. Prilutsky (Georgia Tech.). • Support: Office of Naval Research (ONR) and DARPA.

4. Areas: Neuroscience, robotics, and control. • Applications: • Rehabilitation after spinal cord injury or stroke. • Design and assessment of transplants. • Intraspinal and cortical neuroprosthetics. • Neurorobotics and biomimetic robot designs. • Human and Health Related Impact: • Spinal cord injury and stroke are debilitating to large numbers of individuals. The former affects primarily younger individuals and the latter affects the older population. Transplant and neuroprosthetic interventions, rehabilitation, and assist devices are important aspects of this health care area, and that are impacted significantly by our research. • Military Impact: • Biomimetic and humanoid robots are an area of interest to DARPA, with applications in urban war fighting, and in particular, in remote urban surveillance. NEURAL CONTROL IN THE SPINAL CORD AND MOTOR CORTEX: MODULARITY AND FORCE-FIELD PRIMITIVES P R O J E C T O N E P A G E R • This project focuses on modularity of control in biological motor systems. The goal is to describe neural circuits and associated motor actions that are organized in the spinal cord and motor cortex as collections of visco-elastic force-field primitives. These visco-elastic force-field primitives form a basis to construct both reflexes and novel learned motions and interactions. • Dr. Giszter has a Ph.D. from the Institute of Neuroscience at the University of Oregon. He did postdoctoral research at UCLA and MIT and was then a research scientist at MIT for four years prior to joining the faculty in Neurobiology and Anatomy at MCPHU. • Faculty/Contact: Simon Giszter, Ph.D., Associate Professor in Neurobiology and Anatomy at MCPHU. • E-mail: simon@swampthing.neurobio.mcphu.edu • Collaborators: Michel Lemay, Marion Murray, Alan Tessler, and Itzhak Fischer (MCPHU); John Chapin (SUNY, Brooklyn); Karen Moxon and Ilya Rybak (Drexel University).

5. Neural Prostheses are the Sum of • Electrode technology (metal, solid state, thin-film,      pharmalogical) • Microelectronics • Control algorithms • Biocompatible packaging cortical recording processing spinal injury • Economical Impact   By developing therapies to restore function to   those with existing spinal cord injuries (SCI), the   US would save $10 billion a year for medical and   supportive care alone and as much as $400   billion on future SCI lifetime costs (both direct   and indirect). • Regional Impact   Regional development of the technologies   necessary for clinical applications would bring   about related job opportunities. • Humanity Impact   There are an estimated 250,000 spinal cord   injured individuals living in the US. On average,   11,000 new injuries are reported each year. • Military Impact   Out of the estimated 250,000 Americans living   with spinal cord injury, approximately 40,000 of   these individuals are veterans. The VA has the   largest single network of SCI care in the nation. intraspinal microstimulation INTRASPINAL MICROSTIMULATION FOR MOTOR NEURAL PROSTHESES P R O J E C T O N E P A G E R • The interneuronal circuitry of the spinal cord is capable of generating complex movements with coordinated muscle activity. Current motor neuroprostheses are based on stimulation of the last order neurons, thereby bypassing this organization. Thus, as stated by Burke [1992], "One of the major challenges facing clinical neurobiologists is how to exploit the untapped reserves of coordinated movement contained in the spinal cord circuits of patients with a functionally isolated spinal cord." Intra-spinal micro-stimulation can be used to activate these circuits. A neural prosthesis based on the activation of spinal circuits may improve the control of complex multi-joint movements, and thus provide an increase in functions to individuals with spinal cord injury (SCI). Michel Lemay, Ph.D., is a biomedical engineer with expertise in motor system neural prostheses, biomechanics, robotics, and spinal microstimulation. Dr. Lemay has been working on spinal microstimulation for five years, a period during which he recorded motor system responses (forces, EMGs) evoked by spinal microstimulation in frogs and cats. Prior to this project, Dr. Lemay had experience in clinical biomechanical and fundamental research on motor system neural prostheses • Faculty/Contact: Michel Lemay, Ph.D. (MCP Hahnemann University) • E-mail: Michel.Lemay@drexel.edu • Collaborators: Karen Moxon, PhD. (Drexel University), Simon Giszter, Ph.D. (MCP Hahnemann University), and Warren Grill, PhD. (Case Western Reserve University).

6. T T T Jonathan Nissanov, Ph.D., is a neuro-biologist and biomedical engineer who has worked for many years on the problem of spatial normalization of rodent brain maps. T T INFORMATICS CENTER FOR MOUSE NEUROGENETICS P R O J E C T O N E P A G E R • The purpose of this Neuroinformatics project is to develop and exploit a suite of image databases, motorized Internet microscopes, and software to study the genetic basis of structural variation of the mouse CNS. Resources are open to the research community through an integrated web interface at <nervenet.org>. The focus of this project is to provide a collaborative research environment for mapping quantitative trait loci (QTLs). These genes are responsible for the extraordinary variation in CNS structure among mice and humans. QTL analysis is a burgeoning field that tackles complex biological traits modulated by many genes. We will develop four significant new resources and technologies: The Mouse Brain Library (MBL) consists of a huge, well-organized library of brain sections suitable for morphometric investigation. Thousands of images can be rapidly searched, sorted, and downloaded at a resolution of five microns per pixel using an intuitive and powerful web interface. The Internet Microscope System (iScope) captures and displays extremely detailed movies – Z-axis image stacks – suitable for sophisticated stereological study of all brains in the MBL. The iScope includes robotic slide handlers controlled via the Web 24 hours a day, 7 days a week. The NeuroCartographer Project will develop a suite of software tools and 3D models of hundreds of neuroanatomical structures, enabling researchers to reconstruct and digitally dissect MBL material. The Neurogenetics Tool Box (NTB) comprises a set of gene mapping programs that will enable neuroscientists to rapidly identify and evaluate QTLs responsible for the astonishing variation in CNS architecture. The NTB will include genotypes from an unusually large advanced intercross designed to map loci with sufficient precision, thus enabling a candidate gene approach to cloning QTLs. Achieving the aims of these four projects will catalyze a new era in the structural analysis of the adult mammalian nervous system and will lead to a large number of novel lines of research regarding the development, aging, and pathology of the human brain. • Faculty/Contact: Jonathan Nissanov, Ph.D., Dept. Neurobiology & Anatomy, MCP/Hahnemann University • E-mail: nissanov@drexel.edu • Collaborators: Robert Williams, Dan Goldowitz, and Melburn Park, Dept of Anatomy and Neurobiology, University of Tennessee; Kenneth Manly, Department of Cellular and Molecular Biology, Rosewell Park Cancer Institute; Glenn Rosen, Department of Neurology, Beth Israel Deaconess, Harvard University; Oleh Tretiak, Department of Electrical & Computer Engineering, Drexel University.

7. COMPUTATIONAL MODELING OF NEURAL CONTROL OF BREATHING P R O J E C T O N E P A G E R • The long-term goal of this research is to understand how the brain integrates and coordinates neural processing across multiple levels of organization to produce motor behavior. The neural control of breathing is considered as a model for computational study of cross-level integration of cellular, network, and systems neural mechanisms to provide adaptive motor control. The major objective is to build a comprehensive computational model of neural control of breathing, consistent with experimental data from the cellular to the systems level. • Faculty/Contact: Ilya A. Rybak, Ph.D., Research Professor in BIOMED, Drexel University. • E-mail: rybak@cbis.ece.drexel.edu • Collaborators: Julian F. R. Paton, Ph.D. (Bristol University, UK); Walter M. St. John, Ph.D. (Dartmouth Medical School, NH), Donald R. McCrimmon, Ph.D. (Northwestern University, IL), and Robert F. Rogers, Ph.D. (Thomas Jefferson University, PA). • Support: NSF

8. Applications: • Modeling spinal cord diseases and injuries, as well as methods of their treatment (for the drug discovery industry). • Designing prosthetic neurocontrollers. • Legged robot loco- motion (supported by ONR and DARPA). COMPUTATIONAL MODELING OF THE SPINAL CORD NEURAL CIRCUITRY (NEURAL CONTROL OF LOCOMOTION) P R O J E C T O N E P A G E R • The ultimate goal of this research is to build a comprehensive model of spinal cord neural circuitry that integrates basic reflexes and neuronal central pattern generators for limb control during locomotion. • Faculty/Contact: Ilya A. Rybak, Ph.D., Research Professor in BIOMED, Drexel University. • E-mail: rybak@cbis.ece.drexel.edu • Research interests: Computational neuroscience, neural control, neuro-robotics, and visual perception. • Collaborators: John K. Chapin, Ph.D. (SUNY, NY); M. Antony Lewis (Iguana Robotics, Inc.); Boris I. Prilutsky, Ph.D. (Georgia Institute of Technology, GA). • Web site: http://www.biomed.drexel.edu/faculty/rybak/