Effect of despeckle filtering on classification of breast tumors using ultrasound images

Publication date: Available online 27 February 2019Source: Biocybernetics and Biomedical EngineeringAuthor(s): Kriti, Jitendra Virmani, Ravinder AgarwalAbstractThe present work has been carried out for assessing the performance of texture and morphological features computed from both original and despeckled breast ultrasound images for the classification of breast tumors. Total 100 breast ultrasound images (40 benign and 60 malignant) have been used. Initially, these images have been despeckled using six despeckling filters namely Lee Sigma, BayesShrink, Detail preserving anisotropic diffusion, Fourier ideal, Fourier Butterworth and Homomorphic Fourier Butterworth filter. Total 162 features (149 texture and 13 morphological features) have been computed from both original and despeckled breast ultrasound images for the classification of breast tumors using SVM classifier.Four experiments have been performed in the present work, i.e. (a) analysing the performance of feature set containing texture features and morphological features computed from original images for classification of breast tumors, (b) analysing the performance of feature set containing texture features computed from original images and morphological features computed from despeckled images for classification of breast tumors, (c) analysing the performance of feature set containing texture features computed from despeckled images and morphological features computed from original images for classification of brea...
Source: Biocybernetics and Biomedical Engineering - Category: Biomedical Engineering Source Type: research