Detection of Brain Tumour Using Hybrid Features and Random Forest

Brain tumour is a peculiar entity of mass. Initially, it is a group of few cells but then expands at an accelerating rate. It may originate in the brain or in its surroundings. On a high level, the types of tumour can be benign or malignant. Benign types of tumour are in the initial stages and are curable, whereas the development of malignant types of tumour happens from the glial cells. With the help of various feature extraction techniques, machine learning can be incorporated to detect brain tumour. This paper introduces to application of some efficient feature extraction techniques for brain tumour detection using machine learning. For better results, hybrid features can also be extracted, which is received on combination of techniques. In the proposed model, the best results have been received by using Gabor Filters along with Local Binary Patterns. These techniques have been applied on four machine learning classifiers, i.e. k-Nearest Neighbour, Support Vector Machine, Logistic Regression and Random Forest. The performance of Random Forest is quite encouraging compared to the others.