Automatic Brain Tumor Segmentation using Convolutional Neural Networks with Test-Time Augmentation

TL;DR

This paper explores how applying augmentation techniques during test time, in addition to training, enhances the accuracy and robustness of CNN-based brain tumor segmentation, providing better diagnostic tools.

Contribution

It demonstrates the effectiveness of test-time augmentation in improving CNN segmentation performance and uncertainty estimation for brain tumors.

Findings

Test-time augmentation improves segmentation accuracy.

Augmentation methods include rotation, flipping, scaling, noise.

Enhanced uncertainty estimation for predictions.

Abstract

Automatic brain tumor segmentation plays an important role for diagnosis, surgical planning and treatment assessment of brain tumors. Deep convolutional neural networks (CNNs) have been widely used for this task. Due to the relatively small data set for training, data augmentation at training time has been commonly used for better performance of CNNs. Recent works also demonstrated the usefulness of using augmentation at test time, in addition to training time, for achieving more robust predictions. We investigate how test-time augmentation can improve CNNs' performance for brain tumor segmentation. We used different underpinning network structures and augmented the image by 3D rotation, flipping, scaling and adding random noise at both training and test time. Experiments with BraTS 2018 training and validation set show that test-time augmentation helps to improve the brain tumor segmentation accuracy and obtain uncertainty estimation of the segmentation results.