Offload to high performance devices

Medical monitoring of patients in real-time use applications such as model predictive controllers (MPC). Using a sample taken from the patient, the MPC use a model of human physiology to estimate drug level in parts of patients body to alarm or change drug infusion rate. These computationally heavy MPC models need to estimate the infusion rate every few seconds. In the scenario of monitoring several patient in a hospital using these control applications need to estimate the infusion rate within strict time constraints. This project evaluates the performance of medical control application on many core  platforms like GPUs or Intel Xeon Phi (MIC) or IntelCore i7.

 MPC models in medical control systems for monitoring patients. Using many core devices to meet real-time constraints of these applications.


Researchers: Madhurima Pore, Dr. Ayan Banerjee, Dr. Sandeep K. S. Gupta