MRecover: A Conditional Generative Model for Recovering Motion-Corrupted MR images Using AI Generated Contrast

TL;DR

MRecover is a conditional generative model that synthesizes high-quality TSE MRI images from T1w images to recover motion-corrupted scans, improving data usability and diagnostic analysis in hippocampal studies.

Contribution

The paper introduces MRecover, a novel AI-based model that synthesizes TSE images from T1w images, enhancing motion artifact recovery and data analysis in MRI.

Findings

Achieved high fidelity in synthesized images (SSIM=0.84, FSIM=0.94).

Increased analyzable subjects by 31.8% in a motion-affected dataset.

Enhanced diagnostic effect sizes by increasing sample size.

Abstract

Hippocampal subfield segmentation requires high-resolution T2w turbo spin echo (TSE) MRI, yet this sequence is susceptible to motion artifacts, leading to substantial data loss. We developed a conditional generative model (MRecover) that synthesizes routinely acquired T1w images to create TSE images with autoregressive slice conditioning for volumetric consistency. Trained on 7T MRI data (n=577), the model achieved high in-domain fidelity (n=148, SSIM=0.84, FSIM=0.94) and generalized well to out-of-domain 3T data: subfield volumes from synthesized and the as-acquired images closely matched: (n=416, r=0.87-0.97) and yielded 31.8% more analyzable subjects in the motion-affected ADNI3 dataset after quality control (593 vs 450). The synthesized images also achieved larger effect sizes due to increasing the sample size for diagnostic group differences in hippocampal subfield atrophy (whole hippocampus $\epsilon^2$= 0.121-0.100 vs. 0.086-0.062, left-right hemispheres). Project page: https://jinghangli98.github.io/MRecover/