Analysis and Classification of Renal Ultrasound Images

Abstract

The present research work has been carried out with an aim to enhance the diagnostic potential of conventional B-Mode ultrasound imaging modality for diagnosis of renal diseases. The Computer aided diagnostic (CAD) systems developed are (1) Support vector machine based CAD system for diagnosis of Normal, Medical renal disease and Cyst ultrasound images, (2) Adaptive neuro-fuzzy classifier based CAD system for diagnosis of medical renal disease using absolute texture features, (3) Adaptive neuro-fuzzy classifier based CAD system for diagnosis of medical renal disease using texture ratio features, (4) Adaptive neuro-fuzzy classifier based CAD system for diagnosis of medical renal disease using combined texture features. It was observed that (a) highest overall classification accuracy of 85.9 % was obtained using combined GLCM mean and range features for classification of normal, medical renal disease and cyst classes (b) highest overall classification accuracy of 95.7 % was obtained using GLCM range combined features for classification of normal and medical renal disease classes.