Saved in:
| Main Authors: | , , , , , , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2605.21669 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866910242616377344 |
|---|---|
| author | Li, Jinghang Santini, Tales Clark, Courtney de Almeida, Bruno Chu, Cong Alkhateeb, Salem Sajewski, Andrea Berardinelli, Jacob Jin, Hecheng Campos, Tobias Berardo, Jeremy J. Mettenburg, Joseph Gildengers, Ariel Aizenstein, Howard J. Wu, Minjie Ibrahim, Tamer S. |
| author_facet | Li, Jinghang Santini, Tales Clark, Courtney de Almeida, Bruno Chu, Cong Alkhateeb, Salem Sajewski, Andrea Berardinelli, Jacob Jin, Hecheng Campos, Tobias Berardo, Jeremy J. Mettenburg, Joseph Gildengers, Ariel Aizenstein, Howard J. Wu, Minjie Ibrahim, Tamer S. |
| contents | 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 $ε^2$= 0.121-0.100 vs. 0.086-0.062, left-right hemispheres). Project page: https://jinghangli98.github.io/MRecover/ |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2605_21669 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | MRecover: A Conditional Generative Model for Recovering Motion-Corrupted MR images Using AI Generated Contrast Li, Jinghang Santini, Tales Clark, Courtney de Almeida, Bruno Chu, Cong Alkhateeb, Salem Sajewski, Andrea Berardinelli, Jacob Jin, Hecheng Campos, Tobias Berardo, Jeremy J. Mettenburg, Joseph Gildengers, Ariel Aizenstein, Howard J. Wu, Minjie Ibrahim, Tamer S. Computer Vision and Pattern Recognition Artificial Intelligence 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 $ε^2$= 0.121-0.100 vs. 0.086-0.062, left-right hemispheres). Project page: https://jinghangli98.github.io/MRecover/ |
| title | MRecover: A Conditional Generative Model for Recovering Motion-Corrupted MR images Using AI Generated Contrast |
| topic | Computer Vision and Pattern Recognition Artificial Intelligence |
| url | https://arxiv.org/abs/2605.21669 |