Brain Tumor Analysis using Machine Learning Techniques

Abstract

Abnormal cell growth in brain causes brain tumor which may be a cancerous or non-cancerous one. Tumors can start in the brain, or cancer elsewhere in the body can spread to the brain. This tumor has to be detected at an early stage when it is small. Tumor can be detected easily with the help of MRI modality as the resolution and quality of an image is superior in MRI liken with other imagining techniques and also it is not harmful for the tissues that are present in the brain. In our proposed methodology, three sections have been implemented which includes segmentation, feature extraction and classification. Employment of region based active contour algorithm is used to identify the type of tumors as malignant and benign. Thus, proposed system classifies tumor with an accuracy of 93.34 %.