Brain Tumor Detection using CNN
A classifier based on a convolution neural network (CNN) is considered and Experimental result shows that CNN provides 97.33% accuracy in diagnosis of tumors.
Abstract
- Tumor is an abnormal growth of cell in human brain which is fatal if treatment is not provided at right time. Various imaging technique such as MRI, ultrasound are used to detect the tumor. Earlier, initial stage of tumor were detected manually through observation of image by doctors and the results obtained were sometimes inaccurate and process was time consuming due to large volume of data generated by MRI Scan. Automatic classifications methods are required for proper diagnosis of tumors which can be provided by image processing techniques. Image processing techniques involves four stages such as feature extraction, image segmentation, image pre-processing. In this paper a classifier based on a convolution neural network (CNN) is considered. Experimental result shows that CNN provides 97.33% accuracy.