CATCH: Counterfactual Anatomical Tissue Inpainting with Conditional Haar Diffusion
Quick summary
arXiv:2609.29591v1 Announce Type: cross Abstract: BraTS local synthesis replaces masked regions in T1-weighted brain MRI with plausible tumor-free tissue while preserving observed anatomy. We present CATCH, conditional 3D diffusion in an invertible Haar-wavelet domain. Its denoiser receives noisy target coefficients, voided-image coefficients, and a signed mask; tumor-excluded wavelet reconstruction and a hole-focused loss guide training, and hard compositing preserves observed voxels. We compare fixed masks, tumor-component augmentation, and a weighted mixture of tumor-derived, irregular-blob
Key takeaways
- arXiv:2609.29591v1 Announce Type: cross Abstract: BraTS local synthesis replaces masked regions in T1-weighted brain MRI with plausible tumor-free tissue while preserving observed anatomy.
- We present CATCH, conditional 3D diffusion in an invertible Haar-wavelet domain.
- Its denoiser receives noisy target coefficients, voided-image coefficients, and a signed mask; tumor-excluded wavelet reconstruction and a hole-focused loss guide training, and hard compositing preserves observed voxels.
Why it matters
The importance of “CATCH: Counterfactual Anatomical Tissue Inpainting with Conditional Haar Diffusion” will be measured by what changes in practice. User behavior, access conditions, verifiable performance and responsible-use outcomes are the signals worth following.
Member comments