This paper mainly focuses on the segmentation of MRI images and the creation of a usable image dataset for the tumor detection and VI-RADS scoring mechanism. We described the related works on MRI segmentation and tumor detection algorithms with convolutional neural networks. We explained our approach for segmenting MRI’s using Mask RCNN and giving VI-RADS scores for lesions in bladder wall. The main application area of our project will be on detecting bladder cancers and determining VI-RADS scores using MRI images and deep learning techniques.In this paper, we have put forward a general approach and discussed the tools and methods generally used in MRI handling.