AI Meets HPC: Learning the Cell Motion in Biofluids

Authors: Ziji Zhang, Peng Zhang, and Changnian Han (Stony Brook University); Guojing Cong and Chih-Chieh Yang (IBM - TJ Watson Research Center); and Yuefan Deng (Stony Brook University)

Abstract: 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.

Best Poster Finalist (BP): yes

Poster: PDF
Poster summary: PDF


Back to Poster Archive Listing