双任务步态数据

研究内容：基于本项目所研究的新型认知评估范式，通过多模态同步任务（如EEG（脑电）+fNIRS（近红外脑功能）+眼动+平衡、fNIRS+步态双任务、fNIRS+OT 任务手功能等，其中眼动、平衡、OT 任务手功能为智能传感信息），建立行为特征与神经影像特征的耦合关系；进而优选高敏感性的EEG/fNIRS 通道，构建高准确性的行为学与低通道EEG/fNIRS 的特征组合。在上述优选的基础上，结合认知行为学测试，从感知觉（眼动平衡）检测出发、到执行控制（低通道fNIRS+步态双任务、低通道fNIRS+OT 任务）、记忆与情景（低通道干电极EEG+低通道fNIRS）开展自适应筛查检测，提高筛查的快速性与敏感性。基于大样本数据与深度学习，构建适用于社区的多模态认知障碍筛查模型。针对认知障碍患者在医院的精细分析与康复效果的即时评价的挑战，对照患者/正常人，研究静息态、任务态下的脑活动响应模式，探究认知障碍的神经/血氧响应机制，建立与认知能力相关的EEG/fNIRS 评价指标（包括时空频模式、网络连接等）及确定个体化认知训练靶点。
数据内容：双任务下步态数据；数据采集地点：广东省佛山市；数据采集时间：2020.12-2021.1；设备名称：课题组研发的吉布恩步态检测评估与训练系统；运行环境：无；数据类型：表格；预估数据量/记录数：64例；数据格式：Excel

Research Content: Based on the novel cognitive assessment paradigm investigated in this project, the coupling relationship between behavioral features and neuroimaging features is established through multi-modal synchronous tasks, including EEG (electroencephalogram) + fNIRS (functional near-infrared spectroscopy) + eye movement + balance tasks, fNIRS + gait dual-task, fNIRS + Occupational Therapy (OT) task hand function assessment, etc., where eye movement, balance and OT task hand function data belong to intelligent sensing information. Subsequently, high-sensitivity EEG/fNIRS channels are selected, and feature combinations of high-accuracy behavioral features and low-channel EEG/fNIRS features are constructed. On this basis, combined with cognitive behavioral tests, adaptive screening and detection are carried out, starting from sensory and perceptual detection (eye movement and balance), then executive control detection (low-channel fNIRS + gait dual-task, low-channel fNIRS + OT task), and memory and episodic detection (low-channel dry-electrode EEG + low-channel fNIRS), so as to improve the rapidity and sensitivity of screening. Based on large-sample data and deep learning, a community-applicable multi-modal cognitive impairment screening model is developed. Aiming at the challenges of fine-grained analysis of cognitive impairment patients in hospitals and real-time evaluation of rehabilitation effects, brain activity response patterns under resting state and task state are studied by comparing patients and healthy controls, the neural and hemodynamic response mechanisms of cognitive impairment are explored, EEG/fNIRS-based evaluation indicators related to cognitive ability (including spatiotemporal-frequency patterns, functional network connections, etc.) are established, and individualized cognitive training targets are determined. Data Content: Gait data collected under dual-task conditions; Data collection location: Foshan City, Guangdong Province, China; Data collection time: December 2020 to January 2021; Equipment name: Jibuen Gait Detection, Assessment and Training System developed by the research group; Operating environment: Not applicable; Data type: Tabular data; Estimated data volume/number of records: 64 cases; Data format: Excel

搜集汇总

背景与挑战

背景概述

该数据集是“双任务步态数据”，专注于老年认知障碍的多模态评估与智能康复研究。数据包含64例双任务下的步态记录，采集于广东省佛山市，使用课题组研发的步态检测系统，格式为Excel，数据量1.01MB。其特点在于结合行为特征与神经影像特征，旨在构建认知障碍筛查模型，支持康复医学和人工智能领域的应用。