A Data Augmentation Method for Fully Automatic Brain Tumor Segmentation

TL;DR

This paper introduces TensorMixup, a novel data augmentation technique for 3D brain tumor segmentation that improves model accuracy by synthesizing new training samples through tensor-based image and label mixing.

Contribution

The paper presents TensorMixup, a new augmentation method that enhances brain tumor segmentation accuracy by generating synthetic training data using tensor-based image and label mixing.

Findings

Achieved Dice scores of 91.32%, 85.67%, and 82.20% on different tumor regions.

TensorMixup improves segmentation performance over baseline methods.

Demonstrated feasibility and effectiveness of the proposed augmentation technique.

Abstract

Automatic segmentation of glioma and its subregions is of great significance for diagnosis, treatment and monitoring of disease. In this paper, an augmentation method, called TensorMixup, was proposed and applied to the three dimensional U-Net architecture for brain tumor segmentation. The main ideas included that first, two image patches with size of 128 in three dimensions were selected according to glioma information of ground truth labels from the magnetic resonance imaging data of any two patients with the same modality. Next, a tensor in which all elements were independently sampled from Beta distribution was used to mix the image patches. Then the tensor was mapped to a matrix which was used to mix the one-hot encoded labels of the above image patches. Therefore, a new image and its one-hot encoded label were synthesized. Finally, the new data was used to train the model which could be used to segment glioma. The experimental results show that the mean accuracy of Dice scores are 91.32%, 85.67%, and 82.20% respectively on the whole tumor, tumor core, and enhancing tumor segmentation, which proves that the proposed TensorMixup is feasible and effective for brain tumor segmentation.