基于全光纤复合传感的帕金森病步态检测与识别技术研究

The reliability and portability of sensors in motion capture system is a key factor limiting the development of gait detection and recognition technology in the clinical differential diagnosis of Parkinson's disease. At present, the insufficient response speed, accuracy and stability of the motion sensor result in the serious limitation of the improvement of gait recognition algorithm's accuracy. Aiming at the bottleneck problem, this project will realize a significant improvement in the performance of gait detection and recognition technology through all-fiber composite sensing, and take into account the portability and the particularity of Parkinson's gait. In order to further explore the mechanism of all fiber composite sensing for gait detection and recognition technology performance improvement potential, this project will take the cascaded micro-nano fiber composite sensor as the entry point. The detection mechanism and characteristics of all-fiber composite sensing will be systematically studied. The realization method and demodulation characteristics of gait detection technology will be discussed. The gait recognition calculation model and verification platform of all fiber composite sensor will be established. Thus the acquisition and fusion mechanism of all-fiber multi-source information and its influence on gait detection rules and recognition will be revealed, and will improve the accuracy and reliability of gait recognition technology in the differential diagnosis of Parkinson's disease. This project will provide a completely new technical approach to improve the performance of motion capture systems. And it will also provide a mechanical support for the improvement of the ability of gait differential diagnosis of major diseases such as Parkinson's disease.

运动捕捉系统中传感器的可靠性和便携性是限制步态检测与识别技术在帕金森病临床鉴别诊断发展的一个关键因素。目前，运动传感器不足的响应速度、精度及稳定性导致步态识别算法准确率的提升幅度严重受限。本项目针对该瓶颈问题，通过全光纤复合传感实现步态检测与识别技术性能的大幅提升，并兼顾便携性和帕金森步态特殊性。为了从机理上深入挖掘全光纤复合传感对步态检测与识别技术性能的提升潜力，本项目以级联型微纳光纤复合传感器作为切入点，系统研究全光纤复合传感的探测机理及特性，探索步态检测技术实现方法及解调特性，建立全光纤复合传感的步态识别计算模型及验证平台，从而揭示全光纤多源复合信息融合的内在机制，并获取其对步态检测规则与识别的影响规律，从根本上提高步态检测与识别技术在帕金森病鉴别诊断中的准确性和可靠性。该项目为运动捕捉系统性能提升提供了全新的技术途径，并将为帕金森病等重大疾病的步态鉴别诊断能力的提升提供机理性支撑。

光纤传感技术已经尝试引入步态检测与识别领域，并应用于提取与识别患病者的步态信息。2017年，Alberto课题组报道了基于光纤的足底压力分布远程监测系统，能够从足底压力分布中提取出与疾病相关的足部信息。2018年Frizera课题组报道了基于聚合物光纤传感器进行步态检测，通过获取鞋底光纤传感器内光信号强度变化来检测步态事件。2011年，隋国荣课题组利用光纤列阵传感器的光信号强度变化，分析研究人体在不同运动过程中为保持动态平衡而产生的步态规律。.本项目在现有基于光纤传感的研究成果基础上，以级联型微纳光纤复合传感为探测手段的全光纤型步态检测与识别特性为模型对象，进一步对步态识别的高性能化和便携化进行研究和改进。重点研究级联型微纳光纤复合传感探测机理及可行性，探讨级联型微纳光纤复合传感的实现方法及解调特性，建立基于级联型微纳光纤复合传感的步态检测与识别计算模型及验证平台，努力从理论基础层面到技术实现层面全方位研究基于级联型微纳光纤复合传感的步态检测与识别的传感器设计、实验制备、平台搭建及性能验证优化等一系列问题，具有重要的理论意义和临床医学应用价值。.通过本次科研项目的探索与研究，先后获取如下关键技术：.（1）获取了适用于帕金森病步态识别的光纤复合传感机理及特性；.（2）实现了基于全光纤传感的帕金森病步态检测技术；.（3）构建了基于光纤传感的帕金森病步态识别算法模型。. 申请人及团队在研究实践相关项目过程中，通过前期充分的调研和预研，提出关键环节问题，形成若干拟解决方案，通过实施与验证，得出最优方案，形成关键环节问题的科学认识。并且，通过与同行专家的交流与团队定期的反思总结，解决了如下关键科学问题：.（1）帕金森病步态识别中光纤复合传感机理及其特性；.（2）帕金森病步态识别中全光纤多源复合信息的获取和融合的内在机制；.（3）基于全光纤复合传感的帕金森病步态识别评估模型构建。