CNN for Brain Tumor Segmentation.

Date of Submission

December 2017

Date of Award

Winter 12-12-2018

Institute Name (Publisher)

Indian Statistical Institute

Document Type

Master's Dissertation

Degree Name

Master of Technology

Subject Name

Computer Science


Machine Intelligence Unit (MIU-Kolkata)


Mitra, Sushmita (MIU-Kolkata; ISI)

Abstract (Summary of the Work)

Among brain tumors, gliomas is the most aggressive and common, leading to a very short life expectancy in their highest grade. MRI (Magnetic Resonance Imaging) is a widely used imaging technique to access such tumors but the amount of data produced by MRI is huge which prevents manual segmentation in a reasonable amount of time. So, automatic and reliable methods are required, but the variation in the structure and location of such tumors makes automatic segmentation a very challenging task. In this report, we have proposed four different methods for extracting patches which can be used to train Convolution Neural Networks (CNN) to do the automatic segmentation of tumor in the HGG (Higher Grade Gliomas) and the LGG (Lower Grade Gliomas) patients. We have also proposed a Convolution Neural Network (CNN) based on Transfer Leaming which does automatic segmentation in a reasonable amount of time with promising results for the LGG patients.


ProQuest Collection ID:

Control Number


Creative Commons License

Creative Commons Attribution 4.0 International License
This work is licensed under a Creative Commons Attribution 4.0 International License.


This document is currently not available here.