five

The assessment of stationary and locomotion physical behavior using a single versus dual wearable accelerometer in children who use a manual wheelchair

收藏
DataCite Commons2026-02-13 更新2026-04-25 收录
下载链接:
https://tandf.figshare.com/articles/dataset/The_assessment_of_stationary_and_locomotion_physical_behavior_using_a_single_versus_dual_wearable_accelerometer_in_children_who_use_a_manual_wheelchair/30441096/1
下载链接
链接失效反馈
官方服务:
资源简介:
Wearable devices such as activity monitors can be used to gain insight into children’s physical behavior. However, children who are unable to walk are often excluded from validation studies, and therefore, robust information about their physical behavior is lacking. Therefore, we studied the criterion validity of a wearable prototype activity monitor (AM-p) in children who use a manual wheelchair with and without the ability to walk. We analyzed the data for both single-sensor (sensor placement on upper arm or ankle) and dual-sensor use. Therefore, we conducted a study with cross-sectional design, assessing 37 children (12 girls) aged 6–19 years (mean 12 years, SD 4.3). Children wore an AM-p on the ankle and upper arm and were filmed while performing an activity protocol in a natural setting. Videos were labeled per 5-second epoch with individual activity labels. Raw data were synchronized with labels. An algorithm was trained, and labels were subdivided into pre-defined activity categories. Overall accuracy and F1 score (harmonic mean of precision and recall) were calculated per activity. We demonstrate that the single ankle-worn AM-p can determine “stationary” behavior with excellent accuracy (>90%) and “locomotion” behavior with moderate to good accuracy (77–80%). “Locomotion” behavior includes active wheelchair use of children, which can assist pediatric physical therapists (PPTs) to assess physical behavior correctly in children who use a manual wheelchair. Exploratory analyses indicate that “locomotion” behavior (dual-sensor use), can be divided into “leg activity” and “active wheelchair use” for children who use a manual wheelchair and have the ability to walk. The single ankle-worn AM-p can determine "stationary" behavior with excellent accuracy and "locomotion" behavior with moderate to good accuracy in children who use a manual wheelchair in daily living. Our findings can assist PPTs to assess physical behavior correctly and tailor individual treatment plans. A single ankle-worn activity monitor can determine and discriminate stationary and locomotion behavior in children with disabilities using a wheelchair. In addition, a dual system indicates a correct determination of stationary, leg activity, and active wheelchair use in children with and without the ability to walk.Assessment of physical behavior of children using a manual wheelchair in their everyday life provides insight in activity patterns which, in turn, can help therapists to engage with the child and family for optimal physical performance.By being able to differentiate between child-specific activities, including active and passive wheelchair use, long-term health outcomes and personalized treatment plans can be supported.This article reports on physical activity research in a population at high risk for sedentary behavior and poor health outcomes but often excluded in research. We, with this, promote equity in rehabilitation. A single ankle-worn activity monitor can determine and discriminate stationary and locomotion behavior in children with disabilities using a wheelchair. In addition, a dual system indicates a correct determination of stationary, leg activity, and active wheelchair use in children with and without the ability to walk. Assessment of physical behavior of children using a manual wheelchair in their everyday life provides insight in activity patterns which, in turn, can help therapists to engage with the child and family for optimal physical performance. By being able to differentiate between child-specific activities, including active and passive wheelchair use, long-term health outcomes and personalized treatment plans can be supported. This article reports on physical activity research in a population at high risk for sedentary behavior and poor health outcomes but often excluded in research. We, with this, promote equity in rehabilitation.

诸如活动监测仪(activity monitors)这类可穿戴设备(wearable devices),可用于深入解析儿童的身体行为表现。然而,无法行走的儿童常被排除在验证研究之外,因此相关身体行为的可靠研究数据仍存在缺口。为此,本研究针对使用手动轮椅(manual wheelchair)且具备/不具备行走能力的儿童,评估了一款可穿戴原型活动监测仪(AM-p)的效标效度(criterion validity)。我们分别针对单传感器(single-sensor,传感器佩戴于上臂或脚踝)与双传感器(dual-sensor)的使用场景展开数据分析。本研究采用横断面设计(cross-sectional design),共纳入37名年龄6~19岁的儿童(其中女孩12名,平均年龄12岁,标准差为4.3)。受试者将该原型活动监测仪分别佩戴于脚踝与上臂,并在自然环境下完成预设活动流程的同时进行录像。视频按每5秒时段(epoch)标注个体活动标签,原始数据与标签完成同步匹配。我们训练了专属算法,并将活动标签划分为预设的活动类别。针对每类活动计算总体准确率与F1分数(精确率与召回率的调和均值)。研究结果显示,佩戴于脚踝的单传感器原型活动监测仪可出色识别“静止”行为(准确率>90%),对“移动”行为的识别准确率为中等至良好(77%~80%)。“移动”行为涵盖儿童主动使用轮椅的场景,这一结果可帮助儿科物理治疗师(pediatric physical therapists, PPTs)准确评估使用手动轮椅儿童的身体行为。探索性分析表明,针对具备行走能力且使用手动轮椅的儿童,采用双传感器时的“移动”行为可进一步划分为“腿部活动”与“主动使用轮椅”两类。在日常场景中,佩戴于脚踝的单传感器原型活动监测仪仍可准确识别使用手动轮椅儿童的“静止”行为与“移动”行为。本研究结果可帮助儿科物理治疗师准确评估身体行为,并制定个体化治疗方案。单传感器脚踝佩戴式活动监测仪可识别并区分使用轮椅的残疾儿童的静止与移动行为。此外,双传感器系统可准确识别具备/不具备行走能力儿童的静止行为、腿部活动与主动使用轮椅行为。对日常使用手动轮椅的儿童开展身体行为评估,可揭示其活动模式,进而帮助治疗师与儿童及其家属沟通,以优化身体表现。通过区分儿童特有的活动类型(包括主动与被动使用轮椅),可为长期健康结局与个性化治疗方案提供支持。本研究针对久坐行为风险高、健康结局不佳但常被排除在研究之外的人群开展了身体活动相关研究,旨在推动康复领域的公平性。单传感器脚踝佩戴式活动监测仪可识别并区分使用轮椅的残疾儿童的静止与移动行为。此外,双传感器系统可准确识别具备/不具备行走能力儿童的静止行为、腿部活动与主动使用轮椅行为。对日常使用手动轮椅的儿童开展身体行为评估,可揭示其活动模式,进而帮助治疗师与儿童及其家属沟通,以优化身体表现。通过区分儿童特有的活动类型(包括主动与被动使用轮椅),可为长期健康结局与个性化治疗方案提供支持。本研究针对久坐行为风险高、健康结局不佳但常被排除在研究之外的人群开展了身体活动相关研究,旨在推动康复领域的公平性。
提供机构:
Taylor & Francis
创建时间:
2025-10-24
二维码
社区交流群
二维码
科研交流群
商业服务