Re-examining the association between region-specific pain recurrence and muscle force strategies in patients with patellofemoral pain via OpenSim and artificial intelligence: a prospective cohort study toward targeted rehabilitation.
Zeyi Zhang, Ting Fan, Jin Wu and 1 others
PMID 41126258WHAT IT FOUND
In young male participants with a history of patellofemoral pain, anterior or posterior patellar pain was linked to weaker gracilis and tibialis anterior; lateral pain was linked to weaker rectus femoris and tensor fascia latae during walking.
Key findings
01Anterior or posterior patellar pain recurrence was associated with lower gracilis, tibialis anterior, and internal oblique strength, plus higher adductor longus and tensor fascia latae strength during walking.
02Medial border patellar pain recurrence was associated with higher rectus femoris, gracilis, gluteus maximus, and adductor longus strength, plus lower semitendinosus strength during walking.
03Lateral border patellar pain recurrence was associated with lower rectus femoris, tensor fascia latae, and gluteus maximus strength, plus higher adductor longus and gracilis strength during walking.
STILL TO COME
How it was doneWhat they foundWhat it means for PTs
Read the rest of this summary
You get three full summaries a month, free, and we do not ask for a card. Search, the TL;DRs and your library stay unlimited either way.
What it does not show
The cohort was young and all male, so the findings may not apply to female or older patients. This was an observational cohort, not a trial of rehabilitation; it cannot show that changing these muscle forces prevents pain recurrence. The machine-learning performance was very high, and the authors acknowledge some overfitting may have occurred. The SHAP thresholds are model reference points, not validated clinical cut-offs. OpenSim treated each muscle as one unit, so proximal or distal muscle-region effects were not examined. Pain recurrence was identified by walking pain on a 10-cm scale above 3, not by imaging or a diagnostic test. Four participants sustained ankle sprains, two incurred knee ligament injuries, and five did not complete for personal reasons.
The easy way to misread this
Do not read the machine-learning patterns or thresholds as proof that these muscles caused pain or as a ready-made training prescription. The study observed associations in a young all-male cohort, the authors say the thresholds are not clinically meaningful cut-offs, and the very high model performance may include overfitting.
Summarised by AI from the full paper, without a clinician reviewing it. Check it against the source before it changes what you do. Read it on PubMed →