Anatomically Consistent Cross-Contrast Super-Resolution of Anisotropic Brain T2w MRI

2026-08-08

AI Summary

Anisotropic T2-weighted brain MRI suffers from low resolution in the coronal and sagittal planes, leading to blurred visualization of fine anatomical structures and hindering reliable three-dimensional analysis. To address this limitation, this work proposes the VIPP-SR framework, which enables zero-shot cross-contrast super-resolution reconstruction without requiring isotropic T2w ground truth. Leveraging high-resolution T1c images as guidance, the method employs a View-Invariant Patch-based Generator (VIP-GAN) combined with a shape-preserving strategy and removal of the deepest skip connections to synthesize T2w estimates across all three orthogonal views. Projection consistency optimization is further introduced to enforce anatomical coherence among multi-view reconstructions. Evaluated on BraTS-MET, the approach improves the Dice coefficient from 0.330 to 0.465 and achieves a zero-shot Dice of 0.563 on BraTS-GLI without fine-tuning, substantially outperforming the original anisotropic images.

Abstract

T2-weighted (T2w) brain MRI provides fluid-sensitive soft-tissue contrast that is important for neuro-oncology and radiotherapy planning. However, T2w scans are acquired with anisotropic voxels and appear blurred or stair-stepped on coronal and sagittal views, which obscures small structures and weakens any downstream 3D analysis. We propose VIPP-SR (View-Independent Patched Projection Super-Resolution), a cross-contrast guided super-resolution framework that restores the inter-plane resolution of an existing anisotropic T2w volume without an isotropic ground-truth T2w. VIPP-SR first trains a view-independent patched generator (VIP-GAN) to learn local T1c-to-T2w anatomical correspondence from high-resolution axial slices. The trained generator is then applied to axial, coronal, and sagittal views of the T1c volume to generate three orthogonal T2w estimates. Shape-preserving patching and deepest-skip removal reduce view-specific shortcuts, thereby constraining the generator to learn patch-local representations and enabling the zero-shot inter-plane transfer. Central to VIPP-SR, a projection-based optimization then enforces anatomical consistency across the three view-specific volumes, fusing them by balancing inter-plane self-consistency against per-view data fidelity. The generator is trained on BraTS-MET and evaluated on both the held-out BraTS-MET testing set and the BraTS-GLI cohort without retraining, assessing the cross-cohort generalizability. The results validate that VIPP-SR improves downstream segmentation over the real anisotropic T2w baseline, raising mean-label Dice from 0.330 to 0.465 on BraTS-MET and, zero-shot, from 0.473 to 0.563 on BraTS-GLI and ablation studies identify inter-plane self-consistency as the main source of the gain.

Problem

• anisotropic MRI
• super-resolution
• T2-weighted MRI
• inter-plane resolution
• 3D analysis

Innovation

• cross-contrast super-resolution
• anatomical consistency
• zero-shot inter-plane transfer
• view-independent generator
• projection-based fusion