Development of fast, distributed computational schemes for full body bio-models and their application to novel action potential block in nerves using ultra-short, high intensity electric pulses
An extremely robust and novel scheme for computing three-dimensional, time-dependent potential distributions in full body bio-models is proposed, which, to the best of our knowledge, is the first of its kind. This simulation scheme has been developed to employ distributed computation resources, to achieve a parallelized numerical implementation for enhanced speed and memory capability. The other features of the numerical bio-model included in this dissertation research, are the ability to incorporate multiple electrodes of varying shapes and arbitrary locations. The parallel numerical tool also allows for user defined, current or potential stimuli as the excitation input. Using the available computation resources at the university, a strong capability for extremely large bio-models was developed. So far a maximum simulation comprised of 6.7 million nodes has been achieved for a "full rat bio-model" with a 1 mm spatial resolution at an average of 30 seconds per iteration.
The ability to compute the resulting potential distribution in a full animal body allows for realistic and accurate studies of bio-responses to electrical stimuli. For example, the voltages computed from the full-body models at various sites and tissue locations could be used to examine the potential for using nanosecond, high-intensity, pulsed electric fields for blocking neural action or action potential (AP) propagation. This would be a novel, localized, and reversible method of controlling neural function without tissue damage. It could potentially be used in "electrically managed pain relief," non-lethal incapacitation, and neural/muscular therapy.
The above concept has quantitatively been evaluated in this dissertation. Specifically, the effects of high-intensity (kilo-Volt), ultra-short (∼100 nanosecond) electrical pulses have been evaluated, and compared with available experimental data. Good agreement with available data is demonstrated. It is also shown that nerve membrane electroporation, brought about by the high-intensity, external pulsing, could indeed be instrumental in halting AP propagation. Simulations based on a modified distributed cable model to represent nerve segments have been used to demonstrate a numerical "proof-of-concept."
0544: Electrical engineering