Development and evaluation of an Explainable Prediction Model for Chronic Kidney Disease Patients based on Ensemble Trees

Abstract

<title>Abstract</title> Chronic Kidney Disease (CKD), which implies premature mortality if diagnosed late, is currently experiencing a globally increasing incidence and high cost to health systems. Artificial Intelligence (AI) allows discovering subtle patterns in CKD indicators to contribute to an early diagnosis. In addition, eXplainable AI (XAI) meets the clinicians’ requirement of understanding AI models’ output when patients’ life is eventually affected by the AI algorithms’ decision. This work presents the development and evaluation of an explainable prediction model that would support clinicians in the early diagnosis of CKD patients. The model development is based on a data management pipeline that detects the optimal combination, in terms of classification performance, of ensemble trees algorithms with features selected. The main contribution of the paper involves an explainability-driven approach that allows selecting the best prediction model maintaining a balance between accuracy and explainability that provide quantitative information about the effect of the features selected on the probability of having CKD. Therefore, the most balanced explainable prediction model implements an extreme gradient boosting classifier using 3 features (hemoglobin, specific gravity, and hypertension) that achieves an accuracy of 99.2% and 97.5% with a 5-fold cross-validation and with new unseen data respectively. In addition, an analysis of the model´s explainability shows that hemoglobin is the most relevant feature that influences the prediction results of the model, followed by specific gravity and hypertension. This small number of features selected results in a reduced cost of the early diagnosis of CKD implying a promising solution for developing countries.

Ähnliche Arbeiten