A multicenter team this month reported knee flexion/valgus RMS error around 4–6° and vertical GRF error about 7% versus force plates for 120–240 fps markerless systems during drop jumps in 42 post-ACL athletes. If those numbers hold outside controlled sites, are you updating RTS cutoffs (e.g., <10% asymmetry on DVJ/CMJ) and what sampling rates and calibration routines have been robust in your clinics?
We piloted 180 fps markerless with 25 post-ACLs and kept the <10% DVJ/CMJ cutoff, but we now require two sessions 48–72 h apart plus a quick fatigue block to confirm it sticks. Tip and caveat: lock exposure/white balance, run a 30 s L‑frame calibration each setup, and in dim gyms either bump to 240 fps (fast shutter) or temporarily accept 10–12% with CIs — “not a magic wand” — since “If those numbers hold outside controlled sites” hinges on lighting and reflective floors. What fps are you targeting?
But i’m sticking with the <10% DVJ/CMJ cutoff, but I take the median of 3 landings and repeat it after a 60‑s split‑squat burner to confirm it holds; 240 fps has been cleaner for sub‑200 ms contacts, and at 120 fps I crank shutter speed and do a quick 5‑s static T‑pose + wand pass each block. @t_gibson23’s two‑session idea is gold — my caveat is we flag 8–12% as a gray zone and recheck in identical conditions (“same shoes, same surface”), measure‑twice‑cut‑once.