Session: Best Research Poster Nominees

AI Meets HPC: Learning the Cell Motion in Biofluids

Wednesday, November 18, 2020

10:00 AM – 10:22 AM ET

Channel: Track 8

Author(s)

Ziji Zhang

Stony Brook University, United States of America
Peng Zhang

Stony Brook University, United States of America
Changnian Han

Stony Brook University, United States of America
Guojing Cong

IBM - TJ Watson Research Center
yorktown heights
CY

Chih-Chieh Yang

IBM - TJ Watson Research Center, United States of America
Yuefan Deng

Stony Brook University, United States of America

We generalized the century-old Jeffery orbits equation, by a novel biomechanics-informed online learning framework using simulation data at atomic resolutions, to a new equation of motion for flowing cells to account for the fluid conditions and the cell deformable structures. To validate, we examined the motions, dominantly rotations, of a human platelet in viscous blood flow at various shear stresses and platelet deformability. With the flow and platelet parameters learned from our framework, the new equation captures motions of the platelet accurately. This learned equation will help reduce greatly the complexity of simulating cells in biofluids and, in the case of platelets, of analyzing blood clot formation.