J Hand Surg Eur Vol. 2025 May 11:17531934251340261. doi: 10.1177/17531934251340261. Online ahead of print.
ABSTRACT
We used an artificial intelligence-based model to estimate finger range of movement from video recordings. The mean bias was 11° (standard deviation 34°), with an interclass correlation coefficient of 0.95, indicating high reproducibility.
PMID:40350584 | DOI:10.1177/17531934251340261