TOPIC 425 - AUTOMATED MOTION EXTRACTION AND ANALYSIS TO QUANTIFY PATIENT PERFORMANCE STATUS

Karnofsky评分（1948）和东部协作组ECOG PS（1982）是患者PS评估的标准，并用作进入临床试验和开始或继续任何介入治疗的关键决定因素。Cansera已经展示了使用RGB-D相机的图像分析来量化PS作为癌症护理中的预测生物标志物。RGB-D摄像机捕获在诊所中执行物理任务的患者，并分析患者运动以预测PS。第一阶段SBIR提案将解决以下技术挑战：1）创建从RGB-D相机获取和提取数据的可重复方法; 2）为开放获取库开发RGB-D图像数据集匿名化策略; 3）为不同用户群体制定要求和设计。该项目的成功完成将使商业应用的数据采集和图像捕获使用RGB-D相机，客观地再现匿名方法， 衡量人的绩效，提供定量解决方案。

The Karnofsky score (1948) and the Eastern Cooperative Group ECOG PS (1982) are the standard for patient PS assessment and used as a key determinant for entry into a clinical trial and the initiation or continuation of any interventional therapy. Cansera has demonstrated using image analysis of footage from RGB-D cameras to quantify PS as a predictive biomarker in cancer care. RGB-D cameras captured patients performing a physical task in a clinic and patient movement is analyzed to predict PS. The phase I SBIR proposal will address the following technical challenges: 1) creating a reproducible method for data acquisition and extraction from RGB-D cameras; 2) developing a strategy for anonymizing RGB-D image dataset for an open access library; and 3) establishing requirements and designing for different user groups. The successful completion of the project will enable commercial application for a reproducible anonymized method for data acquisition and image capture using a RGB-D camera to objectively measure human performance providing a quantitative solution.