Development of Methods for Assisting Diagnosis and Rehabilitation of Neuromuscular Disorders based on Musculoskeletal Dynamics Computation

1.神经肌肉骨骼模型的建立及其正反向动力学算法及其基于神经学知识的改进我们建立了一个具有155个自由度的骨骼和大约1000块肌肉的详细的肌肉骨骼人体模型。我们还模拟了脊髓和肌肉之间的网络。利用躯体反射、肌肉动力学等生理学知识对肌肉紧张度估计算法进行改进。分析结果的可视化和降维及其在辅助诊断和康复中的应用我们开发了软件系统来可视化测量和计算的运动和/或肌肉张力数据。为了便于解释结果，我们还开发了降维算法。非侵入性测量识别方法的发展我们开发了许多方法来识别我们的神经肌肉骨骼模型的参数，包括运动学、惯性、关节刚度、肌肉和神经肌肉网络参数。该方法仅使用运动和肌电的非侵入性测量。非侵入性运动测量新方法的提出我们开发了两个原型系统，它们有望取代传统的使用球形标记的光学运动捕获系统。第一个系统利用网状标记来增加测量点的数量，提高测量过程中的安全性。第二种方法是一种无标记的方法，可以同时获得运动和拓扑信息。

1. Development of a Neuromusculoskeletal Model and its Forward/Inverse Dynamics Alforithms, and their Improvements based on Neurological Knowledge We have developed a detailed musculoskeletal human model with 155 degrees-of-freedom skeleton and approximately 1000 muscles. We also modeled the network between the spinal cord and muscles. Physiological knowledge such as somatic reflex and muscle dynamics were used to improve the algorithms for estimating muscle tensions.2. Visualization and Dimension Reduction of Analysis Results and their Application to Assisting Diagnosis and Rehabilitation We developed software systems to visualize measured and computed motion and/or muscle tension data. We also developed algorithms for dimension reduction fur easy interpretation of the results.3. Development of Identification Methods with Noninvasive Measurements We developed a number of methods to identify the parameters of our neuromusculoskeletal model, including kinematics, inertia, joint stiffness, muscle, and neuromuscular network parameters. The methods only use noninvasive measurements of motion and electromyography.4. Proposal of New Methods for Noninvasive Motion Measurement We developed two prototype systems that are supposed to replace conventional optical motion capture systems using spherical markers. The first system utilizes mesh-shaped markers to increase the number of measurement points and improve the safety during measurements. The second method is a markerless approach that can obtain the information on both motion and topology at the same time.