The design of multi texture feature vector classifiers for the diagnosis of ultrasound liver images

The authors propose multi texture feature vector (MTFV) classifiers to diagnose the ultrasound liver images, since single texture feature vector (i.e. coarseness, orientation, regularity, granularity etc.) from the ultrasound images was not sufficiently enough to classify the characteristics of liver diseases. In the authors´ simulation, they used the Bhattacharyya distance (B-distance) and Hotelling Trace Criterion (HTC) to select the best texture feature vectors for the MTFV classifiers and obtained less classification errors than other methods using single texture feature vector. The proposed MTFV classifiers, which used the texture feature vectors and Bayes decision rule, performed well for the classification of normal, fat and cirrhosis liver