Hybrid Feature Extraction Based Brain Tumor Classification using an Artificial Neural Network

The classification of brain tumor is a critical task to formulate a precise decision of the physicians for providing an accurate treatment of the patient according to their classes. Therefore, the classification accuracy needs to be improved to take the proper decision. In this paper, a hybrid feature extraction method with artificial neural network (ANN) is proposed for improving the classification results of the brain tumor. Firstly, the tumor cell region is segmented by using skull stripping with intensity thresholding and region labeling. After that, the segmented tumor cell region is detected by using the canny algorithm and then the features of the detected tumor cell area are employed as the input of the ANN classification network for classification. The achieved classification accuracy shows better performance comparing with some exiting methods that use different classification algorithms with hybrid or single feature extraction methods.