Brain Tumor Segmentation with deep convolutional network approach

Authors

Thibault Buatois

Type

techreport

Year

2019

Number

1905

Abstract

Gliomas are a category of brain tumors that have different degrees of malignancy, shapes and textures. Manual segmentation by experts is a challenging task because of the heterogeneity of these tumors. Several methods of automated gliomas segmentation have been studied at MICCAI 2018 BraTS Challenge. We want to improve the segmentation results submitted last year by LRDE's team, using VGG architecture. This convolutional neural network, classically used for natural image categorization, has been adapted for medical image segmentation through transfert learning and pseudo-3D techniques. Current improvements notably focus on preprocessing, using morphological operators, and will be submitted to this year's challenge.