A video-based system for 3D gait analysis of elderly

基于视频的老年人 3D 步态分析系统

• 批准号: 90291-2013
• 负责人: Meunier, Jean
• 金额: $ 2.11万
• 依托单位: Université de Montréal
• 依托单位国家: 加拿大
• 项目类别: Discovery Grants Program - Individual
• 财政年份: 2013
• 资助国家: 加拿大
• 起止时间: 2013-01-01 至 2014-12-31
• 项目状态: 已结题
• 来源: https://www.nserc-crsng.gc.ca/ase-oro/Details-Detailles_eng.asp?id=521857
• 关键词: video; based; system; 3D; gait

Canada is experiencing significant changes in its population, which is aging rapidly. This major demographic transformation will have far-reaching repercussions on many aspects of personal, social and national life. For instance, one of the greatest hazards for elderly is accidental falls and 62% of injury-related hospitalizations for seniors are the result of falls. It is therefore essential to innovate in the area of elder healthcare to identify innovative technologies to support the elderly. Interestingly enough, gait abnormalities are now considered among the most consistent predictors of falls. In this context, the general objective of this program of research is to develop a multiple-camera system to identify relevant clinical parameters in elder gait for early fall risk assessment and prevention. While most current gait analysis techniques rely on complex and expensive opto-electronic motion capture systems, this research aims at developing a new low-cost, easy-to-use system for use in a clinical setting. It will be based on color-depth synchronized cameras distributed around a treadmill and will use a shape-from-silhouette methodology for the 3D reconstruction of the subject's volume or shape at each time step of the gait cycle. Then statistical analyzes and machine learning will be carried out in order to capture structural and transitional characteristics of the gait for fall risk assessment. Evaluation in a real clinical environment will be conducted to validate the methodology.

加拿大的人口正在经历重大变化，人口正在迅速老龄化。这一重大的人口变化将对个人、社会和国家生活的许多方面产生深远的影响。例如，老年人面临的最大危险之一是意外跌倒，62% 的老年人因受伤住院治疗都是因跌倒造成的。因此，有必要在老年医疗保健领域进行创新，寻找支持老年人的创新技术。有趣的是，步态异常现在被认为是跌倒最一致的预测因素之一。在这种背景下，该研究项目的总体目标是开发一种多摄像头系统来识别老年人步态的相关临床参数，以进行早期跌倒风险评估和预防。虽然大多数当前的步态分析技术依赖于复杂且昂贵的光电运动捕捉系统，但本研究旨在开发一种新的低成本、易于使用的系统，用于临床环境。它将基于分布在跑步机周围的颜色深度同步相机，并将使用轮廓形状方法在步态周期的每个时间步骤对主体的体积或形状进行 3D 重建。然后将进行统计分析和机器学习，以捕获步态的结构和过渡特征，以进行跌倒风险评估。将在真实的临床环境中进行评估以验证该方法。