A novel framework for detection and classification of brain hemorrhage

Abstract

The proposed work focuses on detecting the correct location and type of the hemorrhage in MR Brain image. The Gradient Recalled Echo MR Images are considered as the input image. Then a region and structure specific Multi level Set evolution algorithm is implemented to segment the hemorrhagic region. An enhanced Local Tetra pattern based feature extraction algorithm is used to extract sharpened tetra features and the features are optimized by applying an enhanced Grey Wolf Optimization algorithm. Finally, a Relevance Vector Machine based Classifier is designed to classify the types of the hemorrhages. The proposed framework is compared with the existing techniques on the scale of accuracy, sensitivity, specificity, precision, Jaccard, Dice and kappa coefficient and proved to be outperforming.