OpenCap在人體動作分析的應用：範域文獻回顧

The application of OpenCap for human movement: A scoping review

無標記動作捕捉技術已經廣泛地應用在人體動作的測量，近年有一些研究針對OpenCap在不同動作與群體的應用進行驗證，本研究旨在透過範域回顧的方式彙整目前研究的主要目的、對象、與動作評估類型，瞭解OpenCap在人體動作分析上的應用價值。目前雖有零星研究驗證其效度，但缺乏整體性的應用範疇整理。本研究遵循文獻回顧的原則，利用PubMed、SPORTDiscus和ScienceDirect三個資料庫搜索截至2025年8月有關OpenCap的英文期刊論文，檢視標題和摘要後有25篇進入全文閱讀，最後有22篇納入分析。結果發現OpenCap的研究數量自2023年起快速增長，應用對象從健康成人擴展至運動員及特殊族群。在步態、下蹲與跳躍等動作的下肢矢狀面參數具中至高度一致性，但在額狀、橫切面角度、上肢高速運動及部分動力學參數上仍存在較高的誤差，且受到攝影角度的影響。針對異常步態與臨床族群的測量誤差與變異性偏高，但仍然可以辨識不對稱性與代償策略。目前透過大規模資料與深度學習技術，標記強化模型的演進不斷地提升OpenCap平台在各種動作的測量準確性，這會拓展其在臨床與運動科學的應用範疇。研究者與實務工作者若在實驗室場域之外，可以利用OpenCap針對個體的特定動作型態進行測量，所提供的運動學和動力學參數是值得信任的。

Markerless motion capture technology has been widely applied for human movement measurement, and recent studies have validated OpenCap for various movements and body segments. This study aimed to systematically review the existing literature to summarize research objectives, participant populations, and types of movement assessments, in order to evaluate the application of OpenCap in human motion analysis. Following standard review procedures, a search of PubMed, SPORTDiscus, and ScienceDirect was conducted for English-language journal articles on OpenCap up to August 2025. After screening titles and abstracts, 25 articles were selected for full-text review, with 22 included in the final analysis. Results showed that research on OpenCap has grown rapidly since 2023, with applications expanding from healthy individuals to athletes and special populations. Sagittal plane parameters of lower-limb movements such as gait, squatting, and jumping showed moderate to high consistency, whereas frontal and transverse plane angles, upper-limb high-speed movements, and some kinetic parameters still exhibited errors and were affected by camera perspective. Despite higher errors and variability in atypical gait and clinical populations, gait asymmetries and compensatory strategies were detected. Advances in large-scale datasets and deep learning have improved the accuracy of marker-enhanced models, enabling OpenCap to reliably measure both simple and complicated movements. These developments broaden its potential applications in clinical and sports science, allowing researchers and practitioners to confidently assess specific movement patterns even outside laboratory settings.