Key points are not available for this paper at this time.
This study aims to develop and validate an automated system for assessing hip joint health using fuzzy algorithms and Kinect sensor technology. The primary research question addresses whether a combination of motion capture technology and fuzzy logic can provide reliable, non-invasive evaluation of hip joint mobility comparable to clinical assessment. We utilized a Kinect sensor to measure the active range of motion (ROM) in six distinct directions for 20 patients. An orthopedic specialist evaluated each participant’s joint health, providing clinical assessments that were paired with the ROM measurements. This data was used to develop a three-phase fuzzy algorithm system capable of analyzing joint movement angles. The system categorizes movements individually before integrating the data to produce a comprehensive health evaluation. Under leave-one-out cross-validation (LOOCV) the fuzzy system achieved 90.0% overall accuracy (balanced accuracy 90.0%), with Cohen’s κ = 0.867 and MAE = 0.10 between defuzzified scores and clinician labels. Permutation testing (1000 permutations) yielded p < 0.001, indicating the observed performance is unlikely to arise by chance. Feature-level Kruskal–Wallis tests showed significant differences across clinician classes for all six ROM features (all p ≤ 0.002). This study demonstrates the feasibility of using Kinect sensor technology and fuzzy logic algorithms for objective hip joint health assessment. The developed system offers a cost-effective, non-invasive tool for clinicians to complement traditional diagnostic methods. This approach has the potential to enhance the efficiency and accuracy of hip joint disorder diagnosis in clinical settings.
Yazdani et al. (Wed,) studied this question.