基于传感器的儿童日常运动活动测量准确性研究
该研究纳入31名有移动障碍的儿童和青少年,在住院康复中心进行半结构化活动回路测试,佩戴惯性传感器并以视频标注为参考标准,评估三种传感器配置及算法对日常运动活动的分类准确性。结果显示姿势检测算法准确率97%,轮椅驱动检测算法96%,步行检测算法93%,表明这些配置和算法能较准确测量日常运动活动,但需在临床外长期测试验证。
要点
- 姿势检测算法准确率97%,轮椅驱动检测96%,步行检测93%。
- 研究在住院康复中心进行,样本为31名能步行或使用手动轮椅的儿童。
- 传感器系统需在临床外长期测量中进一步验证后用于临床和科研。
核心结论
- 三种传感器配置及对应算法能较准确测量移动障碍儿童的日常运动活动。
研究设计与样本
- 研究在住院康复中心进行,31名有移动障碍的儿童和青少年参与,佩戴惯性传感器完成半结构化活动回路,以两名独立研究者标注的视频记录为参考标准。
主要发现
- 姿势检测算法分类准确率为97%,轮椅驱动检测算法为96%,步行检测算法为93%。
关键数值与推荐
- 姿势检测算法估计躺、坐、站持续时间及坐站转换次数,准确率97%。
- 轮椅驱动检测算法检测主动和被动驱动期,准确率96%。
- 步行检测算法检测自由和辅助步行期并估计爬楼梯高度变化,准确率93%。
临床怎么用
- 该传感器系统可在临床环境中用于评估移动障碍儿童的日常运动活动,但用于习惯环境前需进行长期测量验证。
局限性
- 研究仅在住院康复中心进行半结构化活动回路测试,未在临床外长期测量中验证。
以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。
原文
OBJECTIVE: This study aimed to determine the accuracy of 3 sensor configurations and corresponding algorithms deriving clinically relevant outcomes of everyday life motor activities in children undergoing rehabilitation. These outcomes were identified in 2 preceding studies assessing the needs of pediatric rehabilitation. The first algorithm estimates the duration of lying, sitting, and standing positions and the number of sit-to-stand transitions with data from a trunk and a thigh sensor. The second algorithm detects active and passive wheeling periods with data from a wrist and a wheelchair sensor. The third algorithm detects free and assisted walking periods and estimates the covered altitude change during stair climbing with data from a single ankle sensor and a sensor placed on walking aids. DESIGN: The participants performed a semi-structured activity circuit while wearing inertial sensors on both wrists, the sternum, and the thigh and shank of the less-affected side. The circuit included watching a movie, playing, cycling, drinking, and moving around between facilities. Video recordings, which 2 independent researchers labeled, served as reference criteria to determine the algorithms' performance. SETTING: In-patient rehabilitation center. PARTICIPANTS: Thirty-one children and adolescents with mobility impairments who were able to walk or use a manual wheelchair for household distances (N=31). INTERVENTIONS: Not applicable. MAIN OUTCOME MEASURE(S): The algorithms' activity classification accuracies. RESULTS: The activity classification accuracy was 97% for the posture detection algorithm, 96% for the wheeling detection algorithm, and 93% for the walking detection algorithm. CONCLUSION(S): The 3 sensor configurations and corresponding algorithms presented in this study revealed accurate measurements of everyday life motor activities in children with mobility impairments. To follow-up on this promising results, the sensor systems needs to be tested in long-term measurements outside the clinic before using the system to determine the children's motor performance in their habitual environment for clinical and scientific purposes.
引用
- Rast FM, Jucker F, Labruyère R, et al. Accuracy of Sensor-Based Measurement of Clinically Relevant Motor Activities in Daily Life of Children With Mobility Impairments. Arch Phys Med Rehabil. 2023. PMID: 37329967. doi:10.1016/j.apmr.2023.05.015
供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。