Abstract
Automatic assessment of hip osteoarthritis severity using the Kellgren–Lawrence grading scale has received considerably less attention than that for knee osteoarthritis, despite its high clinical relevance. From an epidemiological perspective, the incidence of hip osteoarthritis is comparable to that of the knee. Hip and knee OA is one of the leading causes of global disability. In this study, we use the publicly available X-ray dataset addressing the problem of hip arthrosis. We focus the research on the comparative evaluation of machine learning pipelines based on the radiomic features. We investigate how the image preprocessing, feature selection algorithm and finally classification/regression model influence the final outcomes. We compare those settings across three experimental scenarios using a single dataset: arthrosis severity classification, sex classification, and age regression. The findings show that the input preprocessing does not significantly influence the final outcome. In all evaluated cases, the univariate methods (except Spearman) slightly outperform the others. Finally, it was impossible to determine the best model; however, in all cases, the support vector machines ranked highest.
Links to Paper and Supplementary Materials
Main Paper (Open Access Version): https://papers.miccai.org/miccai-2026-sat/paper/RHEUM_AI_009.pdf
SharedIt Link: Not yet available
SpringerLink (DOI): Not yet available
Supplementary Material: Not Submitted
Link to Open Review
Open Review Page: https://openreview.net/forum?id=3oJaZYwDmZ
BibTex
@InProceedings{NurKar_Advantages_MICCAISAT2026,
author = { Nurzyńska, Karolina AND Strzelecki, Michał AND Piórkowski, Adam AND Kamiński, Paweł},
title = { { Advantages and limitations of applying ML techniques in medical data analysis, illustrated with X-ray imaging } },
booktitle = {Medical Image Computing and Computer Assisted Intervention -- MICCAI 2026 Workshops and Challenges},
year = {2026},
publisher = {Springer Nature Switzerland},
volume = {LNCS 17271},
month = {pending},
page = {pending}
}