Category: Rating Scales
Objective: To evaluate the feasibility of video-based motion analysis for rating resting tremor and finger tapping in Parkinson’s disease.
Background: Unified Parkinson’s Disease Rating Scale (UPDRS) is currently the most widely used system to rate the severity of symptom of Parkinson’s disease (PD). However, the manual rating system is time-consuming, semi-quantitative, and rater-dependent. With evolving technology, automated rating system for PD has been developing mainly by wearable device system. We developed a video-clip based motion analysis for resting tremor (RT) and finger tapping (FT) and analyzed its feasibility by its correlation with manual UPDRS ratings.
Method: The study included the video clips of 55 PD patients (110 arms) at resting state and 39 PD patients (78 arms) performing FT tasks. A certified clinician for UPDRS rated each video clips according to UPDRS part III item RT and FT. All videos were framed with Openpose, a pose-estimation deep-learning algorithm, to detect the coordinates of the patients’ hand movements. The displacement of the hand was analyzed as the motion signal. The signals were band-pass filtered and smoothed. For RT, the maximum amplitude (RTmax) and area under curve (RTauc) of hand displacement were measured. Linear regression between RTmax, RTauc with manual RT score was performed. For FT, the taps per second (FTtps), average amplitude (FTamp), fatiguing in amplitude and frequency were measured. Multiple regression model with backward selection between the four factors and manual FT score was performed.
Results: For resting tremor, there was positive correlation between manual RT score with both RTmax (R2 = 0.49, p < 0.001) and RTauc (R2 = 0.45, p < 0.001). For finger tapping, a regression model with the square of FTtps (beta = -0.01, p < 0.001) and FTamp (beta = -0.96, p < 0.001) had good correlation with the manual FT score (R2 = 0.62, p < 0.001).
Conclusion: The video clip-based motion signals could readily predict the manually rated UPDRS scores. By developing this into automated platform, a better cost and time saving system to assess the severity of PD could be introduced.
To cite this abstract in AMA style:
K.W Park, S.Y Jo, E.J Lee, D.W Kang, J.G Lee, J.S Lee, J.H Chung, S.J Chung. Feasibility of video-based motion analysis to rate the severity of tremor and bradykinesia in Parkinson’s disease [abstract]. Mov Disord. 2020; 35 (suppl 1). https://www.mdsabstracts.org/abstract/feasibility-of-video-based-motion-analysis-to-rate-the-severity-of-tremor-and-bradykinesia-in-parkinsons-disease/. Accessed November 22, 2024.« Back to MDS Virtual Congress 2020
MDS Abstracts - https://www.mdsabstracts.org/abstract/feasibility-of-video-based-motion-analysis-to-rate-the-severity-of-tremor-and-bradykinesia-in-parkinsons-disease/