Machine learning for risk stratification of diabetic foot ulcers using biomarkers - Accepted manuscript version

Aug 4, 2026

Upload by Logic

Development of a Diabetic Foot Ulcer (DFU) can sharply reduce patient health and quality of life, making accurate risk stratification essential for guiding timely diabetes care before an ulcer forms. Existing clinical workflows rely on clinicians manually assigning risk categories from biomarker features measured during routine appointments. This paper reports preliminary feasibility-study outcomes using machine learning techniques to stratify DFU formation risk, emphasizing key biomarkers and the role of patient history for risk classification.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages9

Reading Time23 min

AccessFree to download

Abstract

1 Introduction

Diabetic foot ulcers and risk stratification

Motivation for machine learning

Related work and differences

MARTIN, K., UPADHYAY, A., WIJEKOON, A., WIRATUNGA, N. and MASSIE, S. 2023. Machine learning for risk stratification of diabetic foot ulcers using biomarkers. In Mikyška, J., de Mulatier, C., Paszynski, M., Krzhizhanovskaya, V.V., Dongarra, J.J., Sloot, P. M. (eds.) Computational science: proceedings of the 23rd International conference on computational science 2023 (ICCS 2023): computing at the cutting edge of science (ICCS 2023), 3-5 July 2023, Prague, Czech Republic: [virtual event] . Lecture notes in computer science, 14075. Cham: Springer [online], part III, pages 153-161. Available from: [https://doi.org/10.1007/978-3-031-36024-4_11](https://doi.org/10.1007/978-3-031-36024-4_11) Machine learning for risk stratification of diabetic foot ulcers using biomarkers. MARTIN, K., UPADHYAY, A., WIJEKOON, A., WIRATUNGA, N. and MASSIE, S. 2023 This version of the contribution has been accepted for publication, after peer review (when applicable) but is not the Version of Record and does not reflect post-acceptance improvements, or any corrections. The Version of Record is available online at: [https://doi.org/10.1007/978-3-031-](https://doi.org/10.1007/978-3-031-)[ ](https://doi.org/10.1007/978-3-031-)36024-4 11. Use of this Accepted Version is subject to the publisher's Accepted Manuscript terms of use. Machine Learning for Risk Stratification of Diabetic Foot Ulcers using Biomarkers Kyle Martin [0000−0003−0941−3111], Ashish Upadhyay [0000−0003−0124−8879], Anjana Wijekoon [0000−0003−3848−3100], Nirmalie Wiratunga [0000−0003−4040−2496], Stewart Massie [0000−0002−5278−4009] School of Computing, Robert Gordon University, Aberdeen, Scotland {k.martin3, a.upadhyay, a.wijekoon1, n.wiratunga, [s.massie}@rgu.ac.uk](s.massie}@rgu.ac.uk)[ ](s.massie}@rgu.ac.uk)[https://rgu-repository.worktribe.com/tag/1437329/artificial-intelligence-reasoning-air](https://rgu-repository.worktribe.com/tag/1437329/artificial-intelligence-reasoning-air) Abstract. Development of a Diabetic Foot Ulcer (DFU) causes a sharp decline in a patient’s health and quality of life. The process of risk stratification is crucial for informing the care that a patient should receive to help manage their Diabetes before an ulcer can form. In existing practice, risk stratification is a manual process where a clinician allocates a risk category based on biomarker features captured during routine appointments. We present the preliminary outcomes of a feasibility study on machine learning techniques for risk stratification of DFU formation. Our findings highlight the importance of considering patient history, and allow us to identify biomarkers which are important for risk classification. Keywords: Diabetic Foot Ulceration · Machine Learning · Biomarkers 1 Introduction Diabetic Foot Ulcers (DFUs) are a severe complication of Diabetes Mellitus. It is estimated that 15% of diabetic patients will develop a DFU during their lives [12] . Development of a DFU can cause a sharp decline in health and quality of life, often leading to further infection, amputation and death [10, 3] . Predicting the likelihood of DFU formation is based on risk stratification. Risk stratification is crucial for informing the level and regularity of care that a patient should receive. Improper treatment of a DFU can exacerbate patient condition and lead to further health complications, impacting quality of life and increasing cost of treatment [3] . Given medical knowledge of biological markers which act as patient features, clinicians leverage their domain expertise to manually allocate a risk category which describes the likelihood of developing a DFU [1] . Biological markers (henceforth ‘biomarkers’, as per domain terminology) are physiological features captured during routine medical examinations which contribute to a clinician’s understanding of patient condition and their capability to effectively stratify future risk. For example, concentration of albumin in the blood is a re

FAQs

为什么需要对糖尿病足溃疡进行风险分层？

风险分层用于决定患者应获得的护理水平与频率。若治疗不当，可能加重病情并引发更多健康并发症，影响生活质量并增加治疗成本。

现有临床风险分层通常如何进行？

临床上通常由医生基于常规就诊中获取的生物标志物特征，手动分配风险类别，描述未来发生糖尿病足溃疡的可能性。

本文提出的机器学习方法关注哪些关键信息？

研究强调需要考虑患者病史，并用于识别对风险分类具有重要性的生物标志物，从而实现对DFU形成风险的预测性分层。

Aug 4, 2026

Upload by Logic