Brain tumor detection and classification into different types have been implemented using ROI and SVMs. The tumor can be detected properly through our proposed approach and can be classified into benign and malignant with spreading behavior of the tumor that can be either normal, medium, or severely spread. A mixed approach has been used for Feature extraction that consists of PCA OF DWT, GLCM, and LBP for higher accuracy which has been achieved to 98.0932%. Further SVM and different ML techniques can be considered as future work for increasing the accuracy rate near 100% and calculating the percentage of tumor spread that can help in proceeding the right direction with the treatment. In upcoming days, accuracy can be fully achieved by thorough research with time in the given system. Furthermore, different or updated tumor recognition procedures can be introduced giving various options for recognition with advanced features.
展开▼
