Reverse Imaging for Wide-spectrum Generalization of Cardiac MRI Segmentation

2025-08-28

AI Summary

Cardiac MRI segmentation models suffer from limited generalizability across diverse imaging protocols due to substantial contrast variations among sequences. To address this, we propose a physics-driven inverse imaging framework that maps raw MRI acquisitions into an interpretable spin parameter space—namely proton density (PD), longitudinal relaxation time (T1), and transverse relaxation time (T2)—and leverages diffusion models to learn multimodal prior distributions over these parameters. This enables protocol-agnostic domain adaptation and flexible contrast synthesis. Our method jointly integrates nonlinear inverse problem solving with generative modeling and employs mSASHA-based multiparametric data for spin property estimation and physics-consistent data augmentation. Evaluated on highly heterogeneous clinical MRI sequences, the approach consistently improves segmentation accuracy. It represents the first solution achieving broad-spectrum, interpretable, and target-domain-label-free generalization across MRI acquisition protocols.

Technology Category

• Machine Learning: Multimodal Learning
• Computer Vision: Segmentation
• Intelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• User Modeling, Personalization and Recommendation: Federated recommendation systems and personalization
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Pretrained segmentation models for cardiac magnetic resonance imaging (MRI) struggle to generalize across different imaging sequences due to significant variations in image contrast. These variations arise from changes in imaging protocols, yet the same fundamental spin properties, including proton density, T1, and T2 values, govern all acquired images. With this core principle, we introduce Reverse Imaging, a novel physics-driven method for cardiac MRI data augmentation and domain adaptation to fundamentally solve the generalization problem. Our method reversely infers the underlying spin properties from observed cardiac MRI images, by solving ill-posed nonlinear inverse problems regularized by the prior distribution of spin properties. We acquire this "spin prior" by learning a generative diffusion model from the multiparametric SAturation-recovery single-SHot acquisition sequence (mSASHA) dataset, which offers joint cardiac T1 and T2 maps. Our method enables approximate but meaningful spin-property estimates from MR images, which provide an interpretable "latent variable" that lead to highly flexible image synthesis of arbitrary novel sequences. We show that Reverse Imaging enables highly accurate segmentation across vastly different image contrasts and imaging protocols, realizing wide-spectrum generalization of cardiac MRI segmentation.

Problem

• Addressing poor generalization of cardiac MRI segmentation across sequences
• Solving image contrast variations from different MRI protocols
• Enabling accurate segmentation across diverse imaging contrasts

Innovation

• Physics-driven reverse inference of spin properties
• Generative diffusion model for spin prior acquisition
• Interpretable latent variable for flexible image synthesis