Application of Gray Level Co-Occurrence Matrix (GLCM) for Abdominal Wave Image Classification: A Comparative Study of LVQ, KNN, and SVM

Abstract

Medical image classification is a crucial research area in medical imaging analysis to support clinical diagnosis. In this study, we implemented the Gray Level Co-Occurrence Matrix (GLCM) method to extract texture features from abdominal wave images and enhance classification accuracy. Three machine learning classification methods—Learning Vector Quantization (LVQ), K-Nearest Neighbors (KNN), and Support Vector Machine (SVM)—were employed and compared based on their classification performance. The experimental results show that the KNN method achieved the highest accuracy of 96.83%, followed by SVM with 95.24%, and LVQ with 84.13%. These findings indicate that KNN is the most effective classification method for abdominal wave images among those tested. This study highlights the significance of texture feature extraction using GLCM in improving medical image classification accuracy. The results of this study can contribute to the advancement of digital healthcare technologies, particularly in gastrointestinal disorder detection and digestive health monitoring. Future research should explore hybrid deep learning approaches and larger datasets to further enhance classification accuracy and model robustness.