Validity of visual perceptive computing for static posturography in patients with multiple sclerosis

Abstract

Background: Multiple sclerosis (MS) patients frequently have postural control impairment but quantitative posturography is difficult to perform in clinical care. Recent technology facilitates new posturography approaches. Objective: To evaluate construct validity of visual perceptive computing (VPC) for static posturography to study postural control in MS patients. Methods: A total of 90 MS patients and 59 healthy controls (HCs) performed three stance tests: open, closed and tandem stance. Static posturography was performed using a VPC system with Microsoft Kinect. Clinical assessments included Expanded Disability Status Scale (EDSS), Timed-25-Foot-Walk, Short-Maximum-Speed-Walk and 12-item MS Walking Scale (MSWS-12) questionnaire. Reliability was assessed with intra-class correlation coefficients at retest. Results: As a group, MS patients performed worse than HCs in all tests. The closed stance test showed best applicability and reliability. With closed eyes, in 36.7% of patients, the three-dimensional mean angular sway velocity (MSV-3D) was above HCs’ 95th percentile. Higher MSV-3D was associated with decreased walking speed (p < 0.001); worse clinical scores, mainly attributable to the cerebellar functional system score (p < 0.001); and reflected in self-reported walking disability (MSWS-12, p < 0.001). Conclusion: Postural control can be reliably assessed by VPC-based static posturography in patients with MS. Abnormal postural control seems to predominantly reflect involvement of cerebellar circuits with impact on gait and walking disability.