jilab_2021

Our research work focuses on two areas, both are related to the central nervous system: traumatic brain injury and surgical image-guidance in the brain and spine. For the former, our Worcester Head Injury Model (WHIM) integrates information from advanced neuroimaging, vasculature, and latest brain material properties. Its deep learning surrogate generates detailed spatiotemporal strains of the entire brain in near real-time, and in a convenient voxelized format. The rich information provides input to downstream axonal injury models to study cellular axonal damages. The tools enable large-scale and multiscale modeling of traumatic axonal injury to ultimately improve injury detection, mitigation, and prevention.

Here is the Github link for the deep learning or “AI” surrogate of WHIM, highly efficient and highly accurate. Alternatively, google search “CNN brain strain”.

This work has been supported by the NSF (CMMI award #2114697), NIH (R01 NS092853, R21 NS088781), Ford URP (University Research Program), and GPU donation from NVIDIA. Past funding included support from NIH (R21 NS078607), NOCSAE, Hitchcock Foundation, and Neukom Institute, and donation from Intel Inc. We collaborate with colleagues at the University of British Columbia, Virginia Tech, Ford Motor Company, Dartmouth College, Indiana University, Simbex LLC., Stanford, and University of Massachusetts Medical School.

For surgical image-guidance, we translate patient registration techniques previously developed for open skull neurosurgery into applications in open spinal surgery. By registering radiation-free, low-cost intra-operative images such as sterevision and ultrasound with high-quality, pre-operative images such as CT and MRI, image-guidance and navigation can be achieved during surgery. Special registration pipelines are designed to meet the challenges in spine image registration due to intervertebral mobility. Funding for this work is provided by the NIH (R01 EB025747). Past funding support included Dartmouth SYNERGY Scholars Award (NIH K-award) and NIH (R21 NS078607). This work is conducted in collaboration with Dartmouth College, Dartmouth-Hitchcock Medical Center, and Medtronics, Inc.