"A High-Level, One-Slide Research Overview"


The Neural Prosthetic Systems Laboratory (NPSL; Shenoy Group / Shenoy Lab) at Stanford University conducts neuroscience (systems & cognitive neuroscience) and neuroengineering (electrical, bio, and biomedical engineering) research. The group investigates the neural basis of motor preparation and generation, and designs neural prosthetic systems to assist disabled patients. Neural prosthesis research involves desiging, building and testing medical systems which convert electrical signals from neurons in the brain into control signals for prosthetic arms and computer cursors. This central concept is illustrated in the figure above (see brain figure at center). The surrounding figures (a-f) go into a bit more detail, as does the associated figure caption. See Publications for more information.

  • Panels a & b: Front end. Electrical signals are acquired with surgically implanted Bio-MEMS electrode arrays (a, photo from Blackrock Microsystems Inc.), and these signals can be wired out to exterior circuitry. Alternatively, amplification and telemetry circuits (b, photo of "INI3" integrated with a "Utah Array" from Solzbacher/Harrison/Normann Groups at U. Utah) can be integrated on chip to achieve a fully-implanted sensor system. Done in collaboration with Profs. Solzbacher, Harrison & Normann groups at U. Utah and Profs. Meng and Murmann at Stanford.
  • Panels c & d: Signal processing hardware & algorithms. Neural prostheses require accurate and low-power signal processing algorithms for converting neural signals into useful prosthetic control signals (d). These algorithms can be run on light-weight mobile platforms (c), which may eventually be implanted (e.g., included on chip as in b). Done in collaboration with Prof. Meng's group at Stanford.
  • Panel e: System tests. We then simulate and conduct end-to-end prosthetic systems experiments to verify and increase system performance.
  • Panel f: Basic systems neuroscience is central to all aspects of this research, and of considerable basic scientific interest. A deeper understanding of neural function enables the design of higher-performance, more clinically-viable neural prostheses. Our aim is to understand how arm movements are prepared and executed, which might be most clearly understood in terms of dynamical systems theory and low-dimensional state spaces as these methods work toward a "single-trial, millisecond timescale" understanding of neural processing. Done in collaboration with Prof. Sahani at Gatsby/UCL.

Relatively research directions not directly illustrated above include: (1) designing and building an "animal model of freely moving humans" in order to study neural processing in more natural behavioral settings (in collaboration with Prof. Black PhD at Brown University / Max Planck, Profs. Meng PhD and Murmann PhD at Stanford), (2) developing and applying optogenetic techniques in order to perturb neural activity in specific neurons, both for basic neurosience and neural prosthetic applications (in collaboration with Prof. Deisseroth MD PhD at Stanford), and (3) translating cortical neural prostheses to FDA Phase I clinical trial participants (in collaboration with Prof. Henderson MD at Stanford, as part of the Neural Prosthetic Translational Laboratory (NPTL), and BrainGate 2, whose PI is Prof. Leigh Hochberg MD PhD).

Back to NPSL Home Page

Updated: 20 January 2013