Abstract - Maryam Shanechi
Despite successful laboratory demonstrations, recording longevity, performance, generalizability, and robustness remain key challenges hindering clinical viability of motor brainmachine interfaces (BMI). We plan to address these by developing a closed-loopelectrocorticography (ECoG) BMI. We will use ECoG to improve longevity compared to the commonly-used single and multi-unit recordings. One main reason for the key challenges in current BMIs is that they neglect to model the control processes of the brain, which is the controller of movement in any BMI setting. We will build a control-theoretic model of brain behavior during neuroprosthetic control to better infer its intentions. We will construct stochastic observation models of ECoG in various frequency bands to relate it to the brain’s intentions. To adapt to neural non-stationarities and plasticity, we will learn the ECoG model parameters by developing an adaptive estimator that uses the predicted intentions from the control-theoretic model as the teaching signal. We will validate this BMI design in online closed-loop experiments with epilepsy patients implanted with ECoG grids at the USC center for neurorestoration, which has one of the largest patient populations and state-of-the-art electrophysiology equipment. This project will be highly interdisciplinary and conducted at the interface of stochastic control and neuroscience.
AWARDS
| Principal Investigator | Institution | Title | Abstract |
| Andersen, Richard | California Institute of Technology | Engineering Artificial Sensation | View |
| Andrews, Anne | University of California, Los Angeles | Nanoscale Neurotransmitter Sensors | View |
| Bloodgood, Brenda | University of California San Diego | A novel toolkit for visualizing and manipulating activity-induced transcription in living brain. | View |
| Chaumeil, Myriam | University of California, San Francisco | In vivo metabolic imaging of neuroinflammation using hyperpolarized 13C | View |
| Cleary, Michael | University of California, Merced | Capturing physiological maps of neural gene expression | View |
| Cohen, Bruce | University of California, Lawrence Berkeley National Laboratory | Nano-optogenetic control of neuronal firing with targeted nanocrystals | View |
| Dai, Hongjie | Stanford University | Deep brain imaging of single neurons in the second near-infrared optical window | View |
| Hall, Drew | University of California, San Diego | Magnetic Monitoring of Neural Activity using Magnetoresistive Nanosensors | View |
| Krubitzer, Leah | University of California, Davis | An integrated system to monitor, image, and regulate neural activity | View |
| Kubby, Joel | University of California, Santa Cruz | Three-Photon Microscopy with Adaptive Optics for Deep Tissue Brain Activity Imaging | View |
| Melosh, Nicholas | Stanford University | Parallel Solid State Intracellular Patch-Clamping with Biomimetic Probes | View |
| Park, B. Hyle | University of California, Riverside | Label-free 4D optical detection of neural activity | View |
| Portera-Cailliau, Carlos | University of California, Los Angeles | High-speed interrogation of network activity with frequency domain multiplexing | View |
| Shanechi, Maryam | University of Southern California | Control-Theoretic Neuroprosthetic Design Using Electrocorticography Signals | View |
| Smith, Will | University of California, Santa Barbara | Whole brain imaging in a primative chordate | View |
| Wood, Marcelo | University of California, Irvine | Epigenetic PET tracer for cross-species investigation of age-related memory dysfunction | View |