An Analysis of Brain Tumor detection using Machine Learning techniques and Deep Learning Approach

Abstract

Abnormal development of cells in the human body leads to the formation of cancer or tumor. The abnormally formed cells during the cell division are called cancer and they have the property to permeate the nearby tissues of the organs and start affecting the blood and lymphatic system, which is termed as metastasis, thereby reducing the lifespan of the patients. Cancer can have its occurrence in any part of the body and it is categorized mainly from the cell where it originates. Cancer arising from the brain and nervous system is called brain cancer. These brain tumors are classified into benign and malignant. Benign tumors are tumors that have inactive tumor cells and the area of these abnormal regions is structured and can be cured by proper medication. Alternatively, malignant tumors are tumors that have active cells and the area of these abnormal cells that are unstructured cannot be cured by medication. Hence, surgery is required for removing these tumors in the brain image. In conventional methods, brain tumors are detected and diagnosed manually by an expert radiologist. It is a time-consuming and error-prone process. Hence, it is not suitable for high population developing countries. Therefore, computer-aided automatic brain tumor detection and diagnosis methods are preferred. The proposed method of screening using the MR imaging technique is quite simple and fast compared to the traditional methods of screening for brain cancer. This method can also be deployed for a large number of cases quite fast and accurately. Hence this proposed research evolves a technique which involves an MR image of the brain region. It presents a digital imaging system which is able to assist physicians to track brain cancer. The goal is to automatically extract the region where the brain cancer starts to occur. MR imaging techniques are one of the tools to diagnose cancer and to detect and identify the malignant and benign tissues in the human body. In this proposed method Computer-Aided Diagnosis (CAD) is used to diagnose brain cancer. The CAD system for the diagnosis of brain cancer iv requires a segmented brain for the analysis. But most of the previous works had concentrated only on the labelling of the brain and only a few attempts were made to segment the brain automatically from the other anatomical structures. This research work is focused on developing a new automatic segmentation algorithm for segmenting brain tumors from MR images. In this work, five different methods for automatic brain segmentation and classification are proposed.