Intelligent Kidney Abnormality Detection in CT Images Using GLCM and Neural Networks

Kidney stone illness, one of the most common medical problems worldwide, relieves pain and drives people to the ER. Diagnoses often use imaging, but educated doctors must interpret them. Computer-aided diagnostics (CAD) systems help doctors discover kidney stones more accurately. The suggested technique includes preprocessing, feature extraction, and model training. Ultrasound (US) image preprocessing includes restoration, smoothness, sharpness, and contrast enhancement. The Grey Level Co-occurrence Matrices (GLCM) analyze image texture to extract characteristics. Model training uses KELM. The KELM-based model outperforms Extreme Learning Machine (ELM) and K-Nearest Neighbours (KNN) with an average accuracy of 92.54%. This accuracy demonstrates the method can detect kidney stones in ultrasound images. This proposed shows that computer-aided kidney stone detection can be fast and accurate. The combination of GLCM's texture analysis and KELM's classification skills can help medical professionals diagnose kidney stones using imaging.