文献类型：期刊文献

英文题名：A customizable three-dimensional-printed triaxial force sensing finger cot with residual fully connected neural network algorithm for wearable hand rehabilitation

作者：Chen, Yang[1];Bi, Jinglong[1];Hu, Yue[1,2];Gao, Yang[1,2];Hua, Xijin[3];Yan, Yabin[1,2];Jia, Tianyao[4];Xuan, Fuzhen[1,2]

机构：[1]East China Univ Sci & Technol, Sch Mech & Power Engn, Shanghai 200237, Peoples R China;[2]East China Univ Sci & Technol, Key Lab Pressure Syst & Safety, Minist Educ, Shanghai 200237, Peoples R China;[3]Univ Exeter, Dept Engn, Exeter EX4 4QJ, England;[4]Zhejiang Acad Emergency Management Sci & Technol, Hangzhou 310012, Zhejiang, Peoples R China

年份：2026

卷号：530

外文期刊名：CHEMICAL ENGINEERING JOURNAL

收录：;EI(收录号：20260519976997);WOS:【SCI-EXPANDED(收录号:WOS:001678753800001)】；

基金：This research was supported by the National Natural Science Foun-dation of China (Grant Nos. 52275146, 52321002, 61804054, 12411530109 and 12174102) and the Space Application System of China Manned Space Program.

语种：英文

外文关键词：Three-dimensional printing; Electrostatic self-assembly; Triaxial force sensing; Multidimensional force decoupling

摘要：The restoration of hand function is pivotal for regaining autonomy and quality of life following neurological or orthopedic impairments. Multidimensional assessment of finger forces is essential for motor recovery, yet current rehabilitation solutions are often constrained by insufficient multidimensional force sensing capability and poor wearing comfort. Here, we present a real-time monitoring system for assisted hand rehabilitation, integrating a fully customizable triaxial force sensing finger cot with a residual fully connected neural network (Res-FCNN) for force component decoupling. Using high-resolution digital light processing (DLP) three-dimensional (3D) printing, the customizable triaxial force sensing finger cot is fabricated combining complex structures with microscale precision and ergonomic design. The sensitivities of the triaxial force sensor of the system are 0.0213 and 0.0625 N- 1 in tangential and normal directions, respectively. The proposed Res-FCNN model enables precise decoupling of triaxial force components, achieving an R2 greater than 0.95 across all axes. Moreover, experimental validations (including finger motion trajectory distinction and grasp stability assessment) demonstrate that the real-time monitoring system integrating the customizable wearable hardware with algorithm-based signal decoding provides a practical framework for quantitative analysis and evaluation of hand rehabilitation training.