From Force Plates to AI: Establishing Validity and Reliability of a 2D AI-Camera System for Quantifying Sit-to-Stand Power Measurement.

The 24 participants were healthy adults aged 20 to 55, so the results say nothing yet about frail older adults or people whose condition changes how they stand up, such as after joint replacement or with knee osteoarthritis. Testing happened at one clinic with standardised camera placement, lighting and a frame cleared of other people and images. The app cannot tell another person, a photo or a face apart from the person being tested, so behaviour in a busy gym or at home is unknown. Some reference data were lost when the thigh sensor came loose, so only 19 of the 24 people could be compared for validity and 18 for reference reliability. Both repeats were performed in the same session with a seated rest between them, so this is short-term repeatability, not day-to-day reliability. The project was run as quality improvement and was not reviewed by an ethics board. Only one single-camera 2D system was studied, so nothing is known about whether multi-camera 3D systems add anything for out-of-plane movement or asymmetry.

The authors report no professional or personal conflicts of interest and state that the study received no funding, and therefore no financial conflict. The app tested is a commercial product (Exer Scan, Exer Labs, Denver, CO), which the paper notes when describing the equipment but does not address further in its declarations.

Do not treat the camera app and the force plates as interchangeable for the same patient. They agree well on average, but the camera reads higher than the force plates and the difference between the two methods is large enough on its own to look like real change, so choose one method and stay with it for follow-up.