Wei-Ning Lee, PhD | The University of Hong Kong; Jinping Dong, PhD; Haotian Guan, MS
Competition Sponsor: Research Grants Council of the Hong Kong Special Administrative Region, China
Awardee year: 2022
Maintaining functional ability is central to physical and mental well-being in older age. Motility, which is endowed by skeletal muscle, is an intrinsic capacity pertinent to functional ability. Despite a multitude of available measurement tools of muscular function in clinical settings, skeletal muscles exhibit intricate architecture, complex dynamics, individual compositional differences, and adaptive coordination, thus presenting diagnostic discrepancies and therapeutic challenges.
The goal of this project is therefore to develop a personalized data-driven model enabled by a physics-informed neural network (PINN) and noninvasive ultrasonic measures for assessment of muscle quality, which is herein defined by muscle stiffness and movement. More specifically, we will develop a scientific machine learning model, PINN, which is governed by muscle mechanics, instead of treating the model as a black box. This model, coined as mus-PINN, will learn from muscle stiffness and kinematics, which will be quantified by our recently developed ultrasound elastography method. We will evaluate the feasibility of mus-PINN on excised porcine muscle samples and in vivo human calf muscles, which are regarded as the second heart.
This project is foreseen to ultimately 1) provide a new mathematical formulation of the skeletal muscle mechanics in vivo, 2) establish an unprecedented quantitative tool for personalized examination of muscle conditions, and 3) promote a sustainable healthy lifestyle powered by high-quality skeletal muscle as we age.