Dynamic Surgical Prioritization - A Machine Learning and XAI-Based Strategy

Aug 5, 2026

Upload by Rowan

Surgical waiting lists create major operational and equity challenges in healthcare systems, especially when resources are limited and clinical conditions change over time. This work proposes a dynamic, interpretable prioritization framework for surgical patients by combining machine learning, stochastic simulations, and explainable AI. LightGBM predicts temporal prioritization scores qp(t), while simulations capture evolving variables and competitive effects. SHAP explanations provide global and patient-level interpretability, and results on 205 ENT patients show accurate qp(t) estimation and reduced waiting times by up to 26%.

LanguageEnglish

CategoryHealthcare

FormatPDF

Pages20

Reading Time50 min

AccessFree to download

Introduction

Motivation and challenges of surgical waiting lists

Proposed dynamic ML + stochastic simulation + XAI framework

Methodology and modeling approach

LightGBM prediction of qp(t)

Stochastic simulations for temporal evolution

SHAP for interpretability

Article Dynamic Surgical Prioritization: A Machine Learning and XAI-Based Strategy Fabián Silva-Aravena 1, *, Jenny Morales 1, Manoj Jayabalan 2, Muhammad Ehsan Rana 3 and Jimmy H. Gutiérrez-Bahamondes 4 Academic Editors: Xavier Fernando, Pratheepa Jeganathan and Nades Palaniyar Received: 16 December 2024 Revised: 8 January 2025 Accepted: 29 January 2025 Published: 8 February 2025 Citation: Silva-Aravena, F.; Morales, J.; Jayabalan, M.; Rana, M.E.; Gutiérrez-Bahamondes, J.H. Dynamic Surgical Prioritization: A Machine Learning and XAI-Based Strategy. Technologies 2025, 13, 72 . [https://doi.org/10.3390/](https://doi.org/10.3390/)[ ](https://doi.org/10.3390/)technologies13020072 Copyright: © 2025 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license ([https://creativecommons.org/](https://creativecommons.org/)[ ](https://creativecommons.org/)[licenses/by/4.0/](licenses/by/4.0/)) . 1 Facultad de Ciencias Sociales y Económicas, Universidad Católica del Maule, Avenida San Miguel 3605, Talca 3460000, Chile; [jmoralesb@ucm.cl](jmoralesb@ucm.cl) 2 School of Computer Science and Mathematics, Liverpool John Moores University, Liverpool L3 3AF, UK; [m.jayabalan@ljmu.ac.uk](m.jayabalan@ljmu.ac.uk) 3 Technology Park Malaysia (TPM), Asia Pacific University of Technology and Innovation (APU), Kuala Lumpur 50603, Malaysia; [muhd_ehsanrana@apu.edu.my](muhd_ehsanrana@apu.edu.my) 4 Departamento de Ciencias de la Ingeniería, Facultad de Ingeniería, Universidad de Talca, Camino Los Niches Km 1, Curicó 3340000, Chile; [jgutierrezb@utalca.cl](jgutierrezb@utalca.cl) * Correspondence: [fasilva@ucm.cl](fasilva@ucm.cl) Abstract: Surgical waiting lists present significant challenges to healthcare systems, particularly in resource-constrained settings where equitable prioritization and efficient resource allocation are critical. We aim to address these issues by developing a novel, dynamic, and interpretable framework for prioritizing surgical patients. Our methodology integrates machine learning (ML), stochastic simulations, and explainable AI (XAI) to capture the temporal evolution of dynamic prioritization scores, qp(t), while ensuring transparency in decision making. Specifically, we employ the Light Gradient Boosting Machine (LightGBM) for predictive modeling, stochastic simulations to account for dynamic variables and competitive interactions, and SHapley Additive Explanations (SHAPs) to interpret model outputs at both the global and patient-specific levels. Our hybrid approach demonstrates strong predictive performance using a dataset of 205 patients from an otorhinolaryngology (ENT) unit of a high-complexity hospital in Chile. The LightGBM model achieved a mean squared error (MSE) of 0.00018 and a coefficient of determination (R2 ) value of 0.96282, underscoring its high accuracy in estimating qp(t) . Stochastic simulations effectively captured temporal changes, illustrating that Patient 1’s q p(t) increased from 0.50 (at t = 0) to 1.026 (at t = 10) due to the significant growth of dynamic variables such as severity and urgency. SHAP analyses identified severity (Sever) as the most influential variable, contributing substantially to qp(t), while non-clinical factors, such as the capacity to participate in family activities (Lfam), exerted a moderating influence. Additionally, our methodology achieves a reduction in waiting times by up to 26%, demonstrating its effectiveness in optimizing surgical prioritization. Finally, our strategy effectively combines adaptability and interpretability, ensuring dynamic and transparent prioritization that aligns with evolving patient needs and resource constraints. Keywords: dynamic prioritization; machine learning in healthcare; explainable AI; surgical waiting lists; stochastic simulation 1. Introduction Surgical waiting lists pose a significant

FAQs

What problem does the paper address in surgical care?

It targets difficulties caused by surgical waiting lists, where dynamic patient conditions and limited resources make equitable prioritization hard.

How does the proposed method compute dynamic prioritization scores?

It uses LightGBM to predict temporal prioritization scores qp(t), while stochastic simulations model how dynamic variables evolve and interact over time.

How is explainability achieved and validated in the strategy?

SHAP is applied to provide both global and patient-specific explanations of model outputs, helping identify the most influential variables for qp(t) and support transparent decision-making.

Aug 5, 2026

Upload by Rowan