Brain Tumour Image Segmentation Using Deep Learning

Brain Tumour Image Segmentation Using Deep Learning

Rajshree Mondal, Chintha Adithi, Pallavi Mandala, Dammu Francis Wilson

RAJSHREE MONDAL (CSE,CMR Technical Campus)

CHINTHA ADITHI(CSE,CMR Technical Campus)

PALLAVI MANDALA(CSE,CMR Technical Campus)

DAMMU FRANCIS WILSON(CSE,CMR Technical Campus)

---------------------------------------------------------------------***---------------------------------------------------------------------

Abstract –

Automated segmentation of  brain  tumour from  multimodal  MR images  is pivotal  for  the analysis and monitoring of disease progression. As gliomas are malignant and heterogeneous, efficient and accurate segmentation  techniques are  used for  the successful  delineation of  tumours into  intra-tumoral classes. Extensively  used  for  biomedical  image segmentation, Convolutional Neural Networks have significantly improved the state-of-the-art accuracy on the  task  of  brain  tumour  segmentation.  In  this  paper,  we  propose  an  ensemble  of  two  segmentation networks: a 3D CNN and a U-Net, in a significant yet straightforward combinative technique that results in better and accurate predictions.

Key Words:  Deep Learning, BraTS, Medical imaging, Segmentation, U-Net, CNN.