Quantifying tics: Best practices and design considerations for video-based tic coding in research.

Tic disorders (TD), including Tourette Syndrome, are characterized by involuntary, repetitive movements and/or vocalizations that can lead to persistent disability and impairment across the lifespan. Existing research demonstrates that video-based behavioral coding (VBBC) methods can be used to reliably quantify tics, enabling a more objective approach to tic measurement above and beyond standardly used TD questionnaires. VBBC is becoming more popular given the ease and ubiquity of obtaining patient videos. However, rigor and reproducibility of this work has been limited by undescribed and unstandardized approaches to using VBBC methods in TD research. The current paper describes "best practices" for VBBC in TD research, which have been tested and refined in our research over the past 15+ years, including considerations for data acquisition, coding implementation, interrater reliability demonstration, and methods reporting. We also address ethical considerations for researchers using this method.

Automated Quantification of Eye Tics Using Computer Vision and Deep Learning Techniques.

BACKGROUND: Tourette syndrome (TS) tics are typically quantified using "paper and pencil" rating scales that are susceptible to factors that adversely impact validity. Video-based methods to more objectively quantify tics have been developed but are challenged by reliance on human raters and procedures that are resource intensive. Computer vision approaches that automate detection of atypical movements may be useful to apply to tic quantification. OBJECTIVE: The current proof-of-concept study applied a computer vision approach to train a supervised deep learning algorithm to detect eye tics in video, the most common tic type in patients with TS. METHODS: Videos (N = 54) of 11 adolescent patients with TS were rigorously coded by trained human raters to identify 1.5-second clips depicting "eye tic events" (N = 1775) and "non-tic events" (N = 3680). Clips were encoded into three-dimensional facial landmarks. Supervised deep learning was applied to processed data using random split and disjoint split regimens to simulate model validity under different conditions. RESULTS: Area under receiver operating characteristic curve was 0.89 for the random split regimen, indicating high accuracy in the algorithm's ability to properly classify eye tic vs. non-eye tic movements. Area under receiver operating characteristic curve was 0.74 for the disjoint split regimen, suggesting that algorithm generalizability is more limited when trained on a small patient sample. CONCLUSIONS: The algorithm was successful in detecting eye tics in unseen validation sets. Automated tic detection from video is a promising approach for tic quantification that may have future utility in TS screening, diagnostics, and treatment outcome measurement. © 2023 The Authors. Movement Disorders published by Wiley Periodicals LLC on behalf of International Parkinson and Movement Disorder Society.