TransBTS: Multimodal Brain Tumor Segmentation Using Transformer presents a novel segmentation framework that effectively incorporates Transformer in 3D CNN for multimodal brain tumor segmentation in MRI. It is the first attempt to exploit Transformer in 3D CNN for 3D MRI Brain Tumor Segmentation in a way that, through CNN, it achieves modeling of local context information and through Transformers, it achieves learning global semantic correlations. The model has shown significant performance in detecting Enhancing tumor region (ET), Regions of the tumor core (TC) and The whole tumor region (WT) in an MRI. The model is trained on the 3D MRI dataset provided by the Brain Tumor Segmentation (BraTS) 2019 challenge, each sample is composed of four modalities of brain MRI scans.
Input variables : Brain MRI
Output Variables : Segmented MRI
Visit Model : arxiv.org
| Model Category | : | Public |
| Date Published | : | March, 2021 |
| Healthcare Domain | : |
Medical Technology
Provider |
| Code | : | github.com |
| Health Risk Management |
| Disease Detection |