Convolution Neural Network Based Enhanced Computerized Technique for Brain Tumour Detection

摘要

Recent years have witnessed how medical imaging is growing these days. Brain Tumour is considered as the world's 10^{th largest fatal human disease. In this perspective, it is a challenging task for the neurosurgeons to reduce the death rates that occur due to brain tumour. For reducing the death rates, it is the primary duty to identify the tumour cells in their early stage. It is very difficult for the tumour cells to be identified with naked eye for by looking to the MR images. In this paper, an automatic cancer cell detection system is proposed to minimize the burden of the neurosurgeons to identify the cancer cells. Here, CNN (Convolution Neural Network) based system with Median filtering and texture feature extraction is proposed and it gives results with 93 % accuracy and at present it has been proved as the better model for brain tumour cells detection.}