Machine learning-based design of shape-morphing structures for bedsore prevention

Abstract: 

Bedsores, also known as pressure ulcers, are formed by a long-term compression of blood vessels owing to immobilized patients’ limited capacity to shift postures in their hospital beds or wheelchairs, which can result in the degradation of body tissue, or necrosis. In this proposal, we aim to develop a shape-morphing structure that changes the patient's body pressure and posture with a single air pump. The Gu Lab at UC Berkeley plans to collaborate with the MecaWet Group, led by Prof. Bico and Prof. Roman at PMMH-ESPCI in Paris, on this project that consists of 1) inverse design of shape-morphing structures using machine learning algorithms (Gu Lab) and 2) validation of the design process via prototypes fabrication and experimentation (MecaWet Group). We foresee broader applications in various forms of medical equipment, such as cushions for wheelchairs and active belts, that could help paraplegic patients with intestinal transit.

Grace Gu

UC PI:
Grace Gu
Mechanical Engineering, UC Berkeley

Jose Bico

France PI:
Jose Bico
Physique et Mécanique des Milieux Hétérogènes, ESPCI-Paris (École Supérieure de Physique et de Chimie Industrielles de la Ville de Paris)

Author: 
Grace Gu
Jose Bico
Publication date: 
July 1, 2022
Publication type: 
Funded Project