Upper Extremity Kinematic Parameters: Reference Ranges Based on Kinect V2

Stroke and other neurological diseases seriously affect the daily activities of patients. Automated assessment not only quantifies outcomes but also customizes the rehabilitation schedule. However, there are few prescriptive parameters range in automatic assessment. This study provided a set of referable kinematic parameters range based on Kinect V2 from eighteen healthy subjects when they performed a set of upper extremity movements, Fugl-Meyer assessment of upper extremity, reaching movement, and hand-to-mouth movement. We firstly computed kinematic parameters of healthy subjects, including the range of motion of shoulder joint, elbow joint, and the trunk, the relative displacement of elbow and wrist joint to local coordinate frame, the movement smoothness, and the inter-joint coordination. Normality and variance of these parameters were analyzed. To verify the feasibility of these parameters in distinguishing patients with different severity, eight stroke patients were recruited and performed the movements. Same kinematic parameters were calculated from stroke patients and were quantified by qualified coefficient (QC) for each patient. The results showed that the range of motion, movement smoothness, and relative displacement followed a normal distribution in healthy subjects for most movements. These parameters in general were related to the severity, i.e., severer the disease is, lower the QC is. It revealed that the parameters could be used to distinguish patients with different severity. The results contributed to automatic motor function assessment and home rehabilitation of neurological patients based on Kinect V2.