International Journal of Electrical and Computer Engineering (IJECE) Vol. 15, No. 4, August 2025, pp. 4148~4159 ISSN: 2088-8708, DOI: 10.11591/ijece.v15i4.pp4148-4159  4148 Journal homepage: http://ijece.iaescore.com Explainable artificial intelligence and feature based technique for the classification of kidney ultrasound images Fizhan Kausar, Ramamurthy Bojan Department of Computer Science, School of Sciences, CHRIST (Deemed to be University), Bangalore, India Article Info ABSTRACT Article history: Received Aug 16, 2024 Revised Apr 22, 2025 Accepted May 24, 2025 Millions of people worldwide are affected by chronic kidney disease (CKD), which is one of the main causes of death. Using machine learning (ML) models, this study attempts to create a computer-aided diagnostic (CAD) system that can autonomously detect chronic kidney disease (CKD) with improved interpretability. An online medical database provided 340 ultrasound images used in this study, which included both normal and abnormal instances. 94 texture and intensity attributes were obtained from these images using Pyrandiomics. Six machine learning methods were used for classification: According to the evaluation results, support vector machine (SVM), decision tree (DT), random forest (RF), k-nearest neighbors (k-NN), XG-Boost, and naïve Bayes (NB) models were considered. Among these models, the random forest model demonstrated the highest accuracy. Explainable artificial intelligence (XAI) methods, namely Shapley additive explanation (SHAP), were utilized to improve model transparency. Clinicians could be assisted in comprehending the reasoning behind the predictions using SHAP analysis, which identifies the most important features impacting the ML model and visualizes the ranking of each individual feature. Keywords: Explainable AI Kidney ultrasound (US) images Machine learning SHAP Texture feature This is an open access article under the CC BY-SA license. Corresponding Author: Fizhan Kausar Department of Computer Science, School of Sciences, CHRIST (Deemed to be University) Bangalore, India Email: fizhan.kausar@res.christuniversity.in 1. INTRODUCTION Chronic kidney disease (CKD) poses a significant global health threat, especially in low-income countries, due to its increasing prevalence and high mortality rates. Chronic diseases accounted for 60% of global deaths in 2005, increasing to 66.7% by 2020 (WHO). Early prediction and monitoring are essential to curtail CKD progression and prevent severe complications. CKD, marked by prolonged kidney damage or reduced function, incurs high medical costs and increases risks of stroke, heart disease, diabetes, and infections. Despite its widespread impact, CKD remains less recognized compared to other chronic illnesses [1]. Thus, this study considers CKD early prediction and monitoring focusing on pediatrics kidney ultrasound images. Ultrasound (US) is favored for identifying vascular irregularities due to its noninvasive approach and lack of ionizing radiation. Specifically, two-dimensional ultrasound (2-D US) is standard for measuring kidney dimensions and morphology of the kidney, despite challenges posed by image quality variations among low-cost and conventional machines [2]. The texture features in US scan images are used to describe the structural characteristics of tissues. These images can detect subtle structural abnormalities, such as cysts, scarring, and changes in tissue texture, which may indicate the early stages of CKD [3]. Several techniques are used to extract textural information