Robust 3D Brain MRI Inpainting with Random Masking Augmentation
This work addresses dataset biases for brain tumor MRI analysis, representing an incremental improvement over prior winning solutions.
The paper tackled the problem of mitigating dataset biases in brain tumor MRI analysis by developing a deep learning framework for 3D inpainting, achieving first place in the BraTS-Inpainting 2025 challenge with SSIM of 0.919±0.088 and PSNR of 26.932±5.057 on the test set.
The ASNR-MICCAI BraTS-Inpainting Challenge was established to mitigate dataset biases that limit deep learning models in the quantitative analysis of brain tumor MRI. This paper details our submission to the 2025 challenge, a novel deep learning framework for synthesizing healthy tissue in 3D scans. The core of our method is a U-Net architecture trained to inpaint synthetically corrupted regions, enhanced with a random masking augmentation strategy to improve generalization. Quantitative evaluation confirmed the efficacy of our approach, yielding an SSIM of 0.873$\pm$0.004, a PSNR of 24.996$\pm$4.694, and an MSE of 0.005$\pm$0.087 on the validation set. On the final online test set, our method achieved an SSIM of 0.919$\pm$0.088, a PSNR of 26.932$\pm$5.057, and an RMSE of 0.052$\pm$0.026. This performance secured first place in the BraTS-Inpainting 2025 challenge and surpassed the winning solutions from the 2023 and 2024 competitions on the official leaderboard.