DRIFT: Difficulty-aware Rectified Flows for Through-plane MRI Super-Resolution
Computer Vision and Pattern Recognition
Summary
The authors address a common problem in MRI scans where images are blurry between slices because of quick but low-quality scanning. They propose a new two-step method called DRIFT that uses an initial guess of the anatomy followed by a careful, physics-informed process to improve image detail without taking too much time. Their method adjusts how hard it works based on how thick the image slices are and ensures the results stay consistent across different thicknesses. Tests show DRIFT gives better-quality images faster than existing methods.
Authors
Yoonseok Choi, Eun-Gyu Ha, Daniel Kim, Mohammed A. Al-masni, Ming-Hsuan Yang, Dong-Hyun Kim
Abstract
Magnetic Resonance Imaging (MRI) is often acquired with anisotropic resolution to reduce scan time, producing stair-step artifacts along the through-plane direction. In through-plane MRI super-resolution, an efficiency-fidelity trade-off arises: feed-forward regressors are fast but oversmooth at large slice-thicknesses, while sampling-based methods improve fidelity at high inference cost. We propose DRIFT, a two-stage thickness-conditioned rectified flow framework for through-plane MRI super-resolution with continuous input slice-thickness. Stage 1 employs an Anatomical Projection Network (APN) to map low-resolution patches to a coarse high-resolution manifold, providing a deterministic anatomical initialization that shortens the residual transport of Stage 2 and stabilizes slice-wise refinement. Stage 2 refines details via rectified flow and introduces a Physics-Aware Difficulty (PAD) metric derived from slice-thickness induced through-plane bandwidth deficit to guide an Adaptive Integration Scheduler (AIS), allocating ODE steps by thickness. A Consistent Endpoint Trajectory Alignment (CETA) loss enforces thickness-consistent reconstructions. Experiments show that DRIFT outperforms super-resolution baselines while reducing inference cost. Code, models, and interactive demos are available at https://yoonseokchoi-ai.github.io/drift-eccv2026/.