Providing an optimal solution to improve the accuracy of brain tumor detection in MRI images

This paper presents a hybrid method for identifying brain tumors in medical pictures that combines watershed, genetic, and support vector machine methods. By using this technique, the brain tumor is accurately identified and the images are correctly segmented. In order to remove noise from the photos, grayscale and median filters are first applied. After the image has been segmented using the watershed technique, genetic traits are investigated. In order to learn the retrieved features and make an accurate diagnosis of brain tumors, the SVM algorithm is finally used. The evaluation findings, which take into account recall, accuracy, and precision, show that the suggested method performs better than traditional algorithms in segmenting and classifying images, with an accuracy of ۹۵% and precision of ۹۷%.

Keywords:

brain tumor detection , image segmentation , watershed algorithm , MRI images