analiza aiqbench
This article focuses on MPFlow, a novel framework utilizing rectified flow and multi-modal guidance to enhance zero-shot MRI reconstruction without retraining generative priors. By leveraging auxiliary structural scans for cross-modal feature alignment, the method significantly reduces hallucinations and accelerates sampling compared to standard diffusion baselines. The core contribution lies in advancing generative modeling techniques specifically for medical imaging applications through improved data consistency and representation learning.
MPFlow: Multi-modal Posterior-Guided Flow Matching for Zero-Shot MRI Reconstruction
arXiv:2603.03710v4 Announce Type: replace-cross Abstract: Zero-shot MRI reconstruction relies on generative priors, but single-modality unconditional priors produce hallucinations under severe ill-posedness. In many clinical workflows, complementary MRI acquisitions (e.g. high-quality structural scans) are routinely available, yet existing reconstruction methods lack mechanisms to leverage this additional information. We propose MPFlow, a zero-shot multi-modal reconstruction framework built on rectified flow that incorporates auxiliary MRI modalities at inference time without retraining the generative prior to improve anatomical fidelity. Cross-modal guidance is enabled by our proposed self-supervised pretraining strategy, Patch-level Multi-modal MR Image Pretraining (PAMRI), which learns shared representations across modalities. Sampling is jointly guided by data consistency and cross-modal feature alignment using pre-trained PAMRI, systematically suppressing intrinsic and extrinsic hallucinations. Extensive experiments on HCP and BraTS show that MPFlow matches diffusion baselines on image quality using only 20% of sampling steps while reducing tumor hallucinations by more than 15% (segmentation dice score). This demonstrates that cross-modal guidance enables more reliable and efficient zero-shot MRI reconstruction.
Abstract:Zero-shot MRI reconstruction relies on generative priors, but single-modality unconditional priors produce hallucinations under severe ill-posedness. In many clinical workflows, complementary MRI acquisitions (e.g. high-quality structural scans) are routinely available, yet existing reconstruction methods lack mechanisms to leverage this additional information. We propose MPFlow, a zero-shot multi-modal reconstruction framework built on rectified flow that incorporates auxiliary MRI modalities at inference time without retraining the generative prior to improve anatomical fidelity. Cross-modal guidance is enabled by our proposed self-supervised pretraining strategy, Patch-level Multi-modal MR Image Pretraining (PAMRI), which learns shared representations across modalities. Sampling is jointly guided by data consistency and cross-modal feature alignment using pre-trained PAMRI, systematically suppressing intrinsic and extrinsic hallucinations. Extensive experiments on HCP and BraTS show that MPFlow matches diffusion baselines on image quality using only 20% of sampling steps while reducing tumor hallucinations by more than 15% (segmentation dice score). This demonstrates that cross-modal guidance enables more reliable and efficient zero-shot MRI reconstruction.
Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
arXiv:2603.03710 [cs.CV]
(or arXiv:2603.03710v4 [cs.CV] for this version)
https://doi.org/10.48550/arXiv.2603.03710 Focus to learn more arXiv-issued DOI via DataCite
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Text integral sub licența deschisă a publicației · 255 cuvinte
Subiecte conexe
Six Principles of Voice Design: Designing for the Ear and Encoding It into Agent Script
Salesforce Blog · 8 oct. 2026
Presentation: Engineering AI for Creativity and Curiosity on Mobile
InfoQ · 21 iul. 2026
Build a Physical AI model factory with NVIDIA Cosmos 3 on SageMaker HyperPod
AWS · 4 sept. 2026
Generate Autonomous Business Insights with AI Agent and MCP Servers
AWS · 29 iul. 2026