Combining CNNs With Transformer for Multimodal 3D MRI Brain Tumor Segmentation With Self-Supervised Pretraining
This work addresses brain tumor segmentation for medical imaging, presenting an incremental improvement over existing methods.
The paper tackles 3D MRI brain tumor segmentation by modifying TransBTS with architectural changes like Squeeze-and-Excitation blocks and combining it with nnU-Net in an ensemble, achieving Dice scores of 0.8496, 0.8698, and 0.9256 for different tumor regions on the BraTS 2021 validation set.
We apply an ensemble of modified TransBTS, nnU-Net, and a combination of both for the segmentation task of the BraTS 2021 challenge. In fact, we change the original architecture of the TransBTS model by adding Squeeze-and-Excitation blocks, an increasing number of CNN layers, replacing positional encoding in Transformer block with a learnable Multilayer Perceptron (MLP) embeddings, which makes Transformer adjustable to any input size during inference. With these modifications, we are able to largely improve TransBTS performance. Inspired by a nnU-Net framework we decided to combine it with our modified TransBTS by changing the architecture inside nnU-Net to our custom model. On the Validation set of BraTS 2021, the ensemble of these approaches achieves 0.8496, 0.8698, 0.9256 Dice score and 15.72, 11.057, 3.374 HD95 for enhancing tumor, tumor core, and whole tumor, correspondingly. Our code is publicly available.