ArticleBMC sports science, medicine & rehabilitation2026
AI assisted photographic posture evaluation: a comparative study of ChatGPT and expert physiotherapists.
Article in BMC sports science, medicine & rehabilitation, 2026. The graph could read no effect estimate from its abstract, so it casts no vote on the map. Cited by 1 paper.
What it found
Each row is one number read from the abstract, on the scale the paper reported it, with its interval. Left of the dashed line favours the treatment, right favours the comparator. Under each row is the sentence it came from. New to these charts? A ten-minute tutorial.
The abstract states no effect estimate the extractor could read, or names no intervention and outcome on the map, so this paper lights no cell and moves no belief. It is still indexed, cited and linked below.
The trial behind it
Trials whose registry record cites this paper, or whose number appears in the abstract. A trial that started after this paper was published is citing it as background, not reporting it.
Neither the registry nor the abstract names a trial number. If this is a trial report, that itself is worth knowing.
Who cites it
1 citing paper in PubMed.
- Generative AI and Large Language Models in Rehabilitation: A Scoping Review.Life (Basel, Switzerland) · 2026Review
Corrections and comments
PubMed lists nothing against this paper. Absence here is not a guarantee, only a check that was made.
Authors and funding
3 authors.
Funding
No grant is acknowledged in the PubMed record.
Abstract
backgroundTraditional visual methods used in posture analysis are limited by subjectivity and variability, whereas computer vision systems provide precise geometric measurements but require technical expertise and specialized resources. Large language models like ChatGPT offer an accessible, low-cost alternative with human-like interpretive abilities and potential educational value. This study aimed to examine the agreement between ChatGPT and expert physiotherapists in standardized photographic posture assessments.
methodsThis cross-sectional comparative study included 39 young adults with at least one visible postural deviation. The sampling frame consisted of individuals preliminarily evaluated during a posture assessment assignment by physiotherapy students. Standardized anterior, posterior, and lateral photographs were taken under controlled conditions. Two physiotherapists independently assessed posture using a seven-region Posture Analysis Form, and the same anonymized photographs were evaluated by ChatGPT-5.1 (multimodal version, web interface, 15-16 November 2025) using a standardized prompt.
resultsInter-rater agreement between physiotherapists was excellent (κ = 0.84-0.94). Agreement between physiotherapists and ChatGPT ranged from moderate to excellent. Trunk (κ = 0.85) and knee alignment (κ = 0.83) showed almost perfect agreement, while head-neck position showed substantial agreement (κ = 0.67). In contrast, scapular (κ = 0.51), shoulder girdle (κ = 0.47), pelvic (κ = 0.47), and foot alignment (κ = 0.43) exhibited moderate agreement. Percentage agreement ranged from 64.10% to 92.31%.
conclusionThis study suggested that ChatGPT shows high agreement with physiotherapists in regions with clearly observable postural deviations, while its reliability decreases in areas requiring finer angular discrimination. The findings further indicate that the model may serve as a supportive tool, particularly for assessing trunk, knee, and head-neck alignment.
Indexed as
Identifiers
What OpenQuestion holds
Registered trials
Read under generation 80e0d062 · epoch 390. Bibliography from PubMed, PubMed Central and OpenAlex; grants from NIH RePORTER; trial links from ClinicalTrials.gov; estimates, votes and beliefs from the OpenQuestion graph.