AC-Diff: Anatomy-Contrast Disentangled Diffusion for Multi-Vendor Liver MRI Harmonization
- 1University of Texas Health Sciences Center at Houston
Abstract
Magnetic resonance imaging (MRI) exhibits substantial scanner and protocol dependent appearance variations, making harmonization challenging without distorting patient specific anatomy, particularly in abdominal imaging. We propose AC-Diff, an anatomy--contrast disentangled diffusion framework that formulates MRI harmonization as a factor-specific generative intervention. Using aligned multi-contrast supervision and cross-patient factor swapping, AC-Diff learns to separate anatomical structure from acquisition-dependent contrast. Unlike conventional image or latent diffusion, AC-Diff restricts stochastic generation to the contrast subspace while the source anatomy bypasses diffusion and is directly reused during reconstruction. Experiments on an in-house multi-vendor cohort and the external Duke Liver MRI dataset demonstrate improved target-domain alignment with strong structural preservation. In downstream liver segmentation, AC-Diff improves Dice from 0.942 for the original inputs to 0.954 for the harmonized images. These results support contrast-specific latent generation as a promising approach to anatomy-preserving MRI harmonization.
Related
- Joint reconstruction of multi-contrast relaxation-diffusion imaging
- ImmuneLens: linking transcriptional states and TCR clonotypes through disentangled multimodal learning
- The Microbiota Dictates Vendor-Derived Differences in a Murine Clostridioides difficile Infection Model
- DECIPHER integrates disentangled representation learning and prototype-based cell-type deconvolution across molecular modalities
- Retrospective harmonization of neuropsychiatric questionnaires in a large multi-site research consortium using expert mappings and semantic textual similarity
BibTeX
@misc{feng2026acdiff,
title = {{AC-Diff: Anatomy-Contrast Disentangled Diffusion for Multi-Vendor Liver MRI Harmonization}},
author = {H. Feng and H. Dang and L. You and H. Wang and X. Zhou},
howpublished = {bioRxiv},
year = {2026},
month = sep,
doi = {10.64898/2026.09.28.755033},
url = {https://www.biorxiv.org/content/10.64898/2026.09.28.755033v1},
note = {CC BY-NC},
}
RIS
TY - UNPB
TI - AC-Diff: Anatomy-Contrast Disentangled Diffusion for Multi-Vendor Liver MRI Harmonization
AU - H. Feng
AU - H. Dang
AU - L. You
AU - H. Wang
AU - X. Zhou
PY - 2026
DA - 2026/09/28
PB - bioRxiv
DO - 10.64898/2026.09.28.755033
UR - https://www.biorxiv.org/content/10.64898/2026.09.28.755033v1
AB - Magnetic resonance imaging (MRI) exhibits substantial scanner and protocol dependent appearance variations, making harmonization challenging without distorting patient specific anatomy, particularly in abdominal imaging. We propose AC-Diff, an anatomy--contrast disentangled diffusion framework that formulates MRI harmonization as a factor-specific generative intervention. Using aligned multi-contrast supervision and cross-patient factor swapping, AC-Diff learns to separate anatomical structure from acquisition-dependent contrast. Unlike conventional image or latent diffusion, AC-Diff restricts stochastic generation to the contrast subspace while the source anatomy bypasses diffusion and is directly reused during reconstruction. Experiments on an in-house multi-vendor cohort and the external Duke Liver MRI dataset demonstrate improved target-domain alignment with strong structural preservation. In downstream liver segmentation, AC-Diff improves Dice from 0.942 for the original inputs to 0.954 for the harmonized images. These results support contrast-specific latent generation as a promising approach to anatomy-preserving MRI harmonization.
ER -