Author ORCID Identifier
Joshua Paul Tayko Verdillo: https://orcid.org/0009-0007-5012-0982
Lily Ann Duran Bautista: https://orcid.org/0000-0003-4170-3864
Abstract
Introduction. The Functional Movement Screen (FMS) is widely used to identify movement deficiencies and potential injury risk. However, its reliance on visual inspection introduces subjectivity, requires specialized training, and necessitates face-to-face assessment. Advances in computer vision, offer opportunities to automate and better objectify functional movement assessment. This exploratory study aimed to evaluate the feasibility of using a MediaPipe-based system, Software-Optimized Movement Assessment (SOMA), to extract joint range of motion (ROM), joint position, and compensatory movement patterns from patient videos during FMS performance.
Methods. SOMA was developed using MediaPipe Pose and MediaPipe Hands libraries integrated with Python-based tools to analyze videos of individuals performing the seven FMS tests. Standardized camera placement, lighting, and video protocols were used. The software automatically detected body landmarks, calculated joint angles and positions, and applied predefined kinematic thresholds to identify compensatory movements and generate FMS scores. Outputs included annotated videos, joint-angle graphs, CSV files, and diagnostic summaries accessible through a web-based interface.
Results. SOMA successfully extracted joint ROM, joint positions, and movement phases across all seven FMS tests. The system was able to identify compensatory movements and distinguish between correct and incorrect performances using automated thresholds. Joint motion graphs and diagnostic summaries provided objective visualization of movement patterns, demonstrating the feasibility of automated FMS assessment using video-based motion analysis.
Discussion. Findings suggest that SOMA can serve as an objective, accessible, and cost-effective tool for FMS assessment, potentially reducing examiner bias and increasing efficiency in both in-person and remote settings. While the system demonstrates promise as an assistive assessment tool, further research is required to establish its validity and reliability relative to clinician-based ocular inspection and gold-standard motion capture systems.
Recommended Citation
Verdillo, J., & Bautista, L. (2025). MediaPipe-based extraction of joint ROM and position from patient videos for the functional movement screen: An exploratory study. Philippine Journal of Physical Therapy, 4(3), 6-22. https://doi.org/10.46409/002.PFST7818
