High performance classifier for brain tumor detection using capsule neural network

Brain tumor is a kind of dangerous disease that continues to spread worldwide. Therefore, early diagnosis and treatments are extremely important in this case, and it is considered the most common form of cancer that can be found in children and adults. Finding out the right type of tumor in the early stages is an important part for developing a specific treatment and diagnosing the patient’s response according to the treatment. In this paper, we propose a Caps Net (Capsule Neural Network) model to avoid the rise of mortality among the brain tumor people and to reduce the time required for accurate diagnosis. The proposed method involves MRI images for classifying the different types of tumors and outperforms CNN in accuracy.