Assessing the accuracy of Wearable sensor-based motion capturing system application in shoulder movement measurements
Abstract
Shoulder disability or dysfunction can negatively affect different aspects of a patient’s life including their daily activities. To record shoulder movement data for addressing its dysfunctionality, valid assessment of motion is important. Wearable technologies have been more recently and offer advantages such as being lightweight, time efficient, usable at home, and user friendly. However, these new technologies have not been evaluated thoroughly in terms of validity and reliability. Combining multiple sensors provides the opportunity to evaluate more complex motions such as those that occur in the upper extremity. A custom motion shirt was designed for this purpose. The objective of this thesis was to validate the accuracy of wearable sensor-based motion shirt. To conduct this study, the performance of the motion shirt was compared to that of the Dartfish tool. First, 10 healthy participants without any shoulder disability aged 50 years old and over with English proficiency were hired. They were asked to perform standard FIT-HaNSA shoulder tasks while they wore motion shirt with active sensors recording the shoulder motions. While doing the tasks, they were also filmed to measure their shoulder movements in Dartfish video analysis software. In a few tasks, the motion shirt sensors did not record the motion data due to lack of continuous sensors synchronization. For 12 tasks that the motion shirt recorded the data successfully, the measured shoulder flexion-extension measures and the calculated shoulder ROFE, (Range of Flexion-Extension) were compared to those measured in Dartfish using Intraclass Coefficient Correlation (ICC) and Bland-Altman plots, respectively. ICC values were over 0.9 in 10 out of 12 tasks and about 0.8 or over in two of the tasks. These values showed a very good correlation between the motion shirt and Dartfish measurements of shoulder extension-flexion measurements in 10 tasks and a good correlation in two tasks. Bland-Altman plots demonstrated that the differences between the shoulder ROFEs (maximum angle subtracted by minimum for each of motions) calculated using motion shirt data and Dartfish analyses were often very similar and maximum 30 degrees different. This proved that motion shirt can be a valid tool for assessing shoulder ROFEs. Results showed that the designed motion shirt was as validate as the Dartfish motion analysis software in quantifying shoulder movements and can be considered a valid tool in this regard. However, further research is necessary to validate these findings and to conduct a more comprehensive assessment.