A Conceptual Framework to Predict Disease Progressions in Patients with Chronic Kidney Disease - Using Machine Learning and Process Mining - Abstract

Aug 2, 2026

Upload by Himbo

Process Mining analyzes existing process flows by mining event logs, while Machine Learning in health data science aims to replicate human decision-making via algorithms. Prior work often applies either Process Mining or Machine Learning separately for healthcare, leaving limited evidence on their joint use. This paper proposes a practical hybrid framework combining Process Mining and Machine Learning to support disease progression prediction for chronic kidney disease. It also evaluates how time-stamped information influences model performance.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages4

Reading Time10 min

AccessFree to download

Introduction

Problem context: CKD progression and clinical need

Electronic Health Records and event logs

Study design: MIMIC-IV and timestamp engineering

Existing framework

Prior work on process mining and learning in healthcare

190 Healthcare Transformation with Informatics and Artificial Intelligence J. Mantas et al. (Eds.) © 2023 The authors andIOS Press. This article is published online with Open Access by IOS Press and distributed under the terms of the Creative Commons Attribution Non-Commercial License 4.0 (CC BY-NC 4.0). doi:10.3233/SHTI230459 A Conceptual Framework to Predict Disease Progressions in Patients with Chronic Kidney Disease, Using Machine Learning and Process Mining Nichalini KANDASAMYa , Thierry CHAUSSALETa,1 and Artie BASUKOSKIaa School of Computer Science & Engineering, University of Westminster, London, UK Abstract. Process Mining is a technique looking into the analysis and mining of existing process flow. On the other hand, Machine Learning is a data science field and a sub-branch of Artificial Intelligence with the main purpose of replicating human behavior through algorithms. The separate application of Process Mining and Machine Learning for healthcare purposes has been widely explored with a various number of published works discussing their use. However, the simultaneous application of Process Mining and Machine Learning algorithms is still a growing field with ongoing studies on its application. This paper proposes a feasible framework where Process Mining and Machine Learning can be used in combination within the healthcare environment. Keywords. Process Mining, GMM algorithm, Hybrid algorithm, DREAM, Chronic Kidney Disease 1. Introduction The study will focus on Chronic Kidney Disease, which 3.5 million in the UK alone.[1] CKD is comorbid with diabetes, hypertension and cardiovascular disease.[2] The disease progresses throughout 5 stages and at various rate depending on the patient’s other medical conditions. With the progression of CKD, patients are at higher risk of developing cardiovascular disease and lower quality of life. An accurate and advanced evaluation of the disease progression can help clinicians and patients to get the most beneficial treatment to slow down the progression of the disease. Early knowledge of possible future diagnoses can guide medical practitioners in administrating an appropriate medical examination and treatment. When treating a patient, their medical details are recorded and stored into a database which can be accessed by clinicians to view and analyze patient records. Those patient records are known as Electronic Health Records (EHR) and have historical medical information about patients that are regularly maintained and updated by the database owner (e.g., hospital, GP practices) . Electronic Health Records are formed of various 1 Corresponding Author: Thierry Chaussalet, E-mail: [chausst@westminster.ac.uk](chausst@westminster.ac.uk) N. Kandasamy et al. / A Conceptual Framework to Predict Disease Progressions in Patients 191 datasets that contain laboratory results, radiographies, clinical notes and observations, progress reports, historical medication records and patients’ personal information. The principal purpose of the EHR is to accurately record patients’ information and medical progression to avoid duplication and inappropriate administration of treatment. [3] Each historical diagnosis is captured with a timestamp and diagnosis code (ICD codes – International Classification of Diseases) that can be mapped onto an event log. The event log can then be mined to understand the existing treatment flow and gain an understanding of the disease progression using historical health records. The process model can help convert timestamp events into a variable that can be used as input variable for the prediction model and determine the next probable diagnosis code. Existing studies mainly focus on either the sole application of process mining or machine learning to predict disease progression. The use of timestamped variables is never considered due to the complexity of engineering those variables for machine learning models. In this study, the MIMIC-

FAQs

Why is predicting chronic kidney disease progression clinically important?

CKD advances through multiple stages and at different rates depending on comorbid conditions. Accurate progression evaluation can help clinicians and patients select beneficial treatment strategies to slow disease advancement and reduce complications.

How do electronic health records support the proposed prediction approach?

EHRs store historical patient data such as lab results, clinical notes, progress reports, and medication histories. Diagnoses with timestamps and ICD codes can be mapped to an event log, enabling mining of treatment flow and disease progression patterns.

What is the role of Process Mining and Machine Learning in the framework?

Process Mining is used to mine event logs and derive insights into existing treatment flow and progression. Machine Learning builds prediction capability, and the combined framework is intended to improve forecasting by integrating timestamped information engineered via the DREAM algorithm.

Aug 2, 2026

Upload by Himbo