Skip to collection list Skip to video grid
Skip to collection list Skip to video grid

All Videos

Accelerated Characterization of Neural Circuits of the Brain

Simon Warfield Benoit Scherrer It is now recognized that a number of brain disorders, including concussion, autism and schizophrenia, are disorders of neural circuitry. However, the routine approach to diagnosis and prognosis in neurology has not changed: it is still focused on assessing patient symptoms instead of assessing neural circuits themselves. Over the last decade, diffusion compartment magnetic resonance imaging has developed characterize the size, the connectivity and the microstructural integrity of specific neural circuits in vivo. Unfortunately, the computation needed currently takes too long for these to be available in routine clinical care. Through our IPCC we have exploited modern parallel computing techniques and optimizations for Intel Xeon & Intel Xeon Phi to dramatically accelerate computation, obtaining ultimately a 160x acceleration (42h->16min). This effectively allows the assessment of the integrity of neural circuits in routine clinical care and will help to facilitate diagnosis, prognosis, and evaluation of the success of therapies. Modern parallel computing techniques and optimizations have allowed dramatic acceleration of computations to characterize neural circuits of the brain.