An Unsupervised Machine Learning Approach for Clustering Hip Arthroplasty Patients: Surgery Duration Differs Among Different Patient Groups

Mohammad Chavosh Nejad, Rikke Vestergaard Matthiesen, Iskra Dukovska-Popovska, John Johansen

2025

Abstract

Operating Rooms (ORs), as the largest source of revenue and costs in hospitals, face the challenge of growing demand while dealing with limited resources, emphasizing the need for operational efficiency. Duration of surgery (DOS), a key element in planning surgical resources, fluctuates and depends on many factors including patients’ characteristics. A better understanding of these factors and the way they affect DOS can help OR planners in achieving efficient resource allocation. To distinguish between patients from the DOS perspective, this paper proposes an unsupervised machine learning method that clusters patients into different groups by considering different clinical and operational features. Seven relevant factors were extracted from Aalborg University Hospital’s database for 1,847 patients undergoing hip arthroplasty. K-Prototype algorithm was utilized for developing various clustering models and their performance was assessed by three popular metrics. Among the different developed models, the one with 7 clusters achieved the highest performance. One-way ANOVA analysis illustrated that DOS means are significantly different among different clusters (F-statistic=11.77, P-Value=5.45e-13). Inter-cluster differences were analyzed by Turkey’s Honest Significant Difference (HSD) test. Besides, evaluating features’ importance showed that Age, BMI, and surgery type are the most contributing factors in clustering patients.

Paper Citation

in Harvard Style

Chavosh Nejad M., Matthiesen R., Dukovska-Popovska I. and Johansen J. (2025). An Unsupervised Machine Learning Approach for Clustering Hip Arthroplasty Patients: Surgery Duration Differs Among Different Patient Groups. In Proceedings of the 18th International Joint Conference on Biomedical Engineering Systems and Technologies - Volume 2: HEALTHINF; ISBN 978-989-758-731-3, SciTePress, pages 398-405. DOI: 10.5220/0013108700003911

in Bibtex Style

@conference{healthinf25,
author={Mohammad Chavosh Nejad and Rikke Matthiesen and Iskra Dukovska-Popovska and John Johansen},
title={An Unsupervised Machine Learning Approach for Clustering Hip Arthroplasty Patients: Surgery Duration Differs Among Different Patient Groups},
booktitle={Proceedings of the 18th International Joint Conference on Biomedical Engineering Systems and Technologies - Volume 2: HEALTHINF},
year={2025},
pages={398-405},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0013108700003911},
isbn={978-989-758-731-3},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 18th International Joint Conference on Biomedical Engineering Systems and Technologies - Volume 2: HEALTHINF
TI - An Unsupervised Machine Learning Approach for Clustering Hip Arthroplasty Patients: Surgery Duration Differs Among Different Patient Groups
SN - 978-989-758-731-3
AU - Chavosh Nejad M.
AU - Matthiesen R.
AU - Dukovska-Popovska I.
AU - Johansen J.
PY - 2025
SP - 398
EP - 405
DO - 10.5220/0013108700003911
PB - SciTePress