Related Experiment Video
Updated: Aug 11, 2026
MRI and PET in Mouse Models of Myocardial Infarction
Published on: December 19, 2013
Translating MRI to PET through conditional diffusion models with enhanced pathology awareness
Yitong Li1, Igor Yakushev2, Dennis M Hedderich3
1Lab for Artificial Intelligence in Medical Imaging, Institute for Diagnostic and Interventional Radiology, School of Medicine and Health, TUM Klinikum, Technical University of Munich (TUM), Munich, Germany; Munich Center for Machine Learning (MCML), Munich, Germany.
Abstract:
Positron emission tomography (PET) is a widely recognized technique for diagnosing neurodegenerative diseases, offering critical functional insights. However, its high costs and radiation exposure hinder its widespread use. In contrast, magnetic resonance imaging (MRI) does not involve such limitations. While MRI also detects neurodegenerative changes, it is less sensitive for diagnosis compared to PET. To overcome such limitations, one approach is to generate synthetic PET from MRI. Recent advances in generative models have paved the way for cross-modality medical image translation; however, existing methods largely emphasize structural preservation while neglecting the critical need for pathology awareness. To address this gap, we propose PASTA, a novel image translation framework built on conditional diffusion models with enhanced pathology awareness. PASTA surpasses state-of-the-art methods by preserving both structural and pathological details through its highly interactive dual-arm architecture and multi-modal condition integration. Additionally, we introduce a novel cycle exchange consistency and volumetric generation strategy that significantly enhances PASTA's ability to produce high-quality 3D PET images. Our qualitative and quantitative results demonstrate the high quality and pathology awareness of the synthesized PET scans. For Alzheimer's diagnosis, the performance of these synthesized scans improves over MRI by 4%, almost reaching the performance of actual PET. Our code is available at https://github.com/ai-med/PASTA.
Insights
We developed PASTA, a new AI framework to create synthetic Positron Emission Tomography (PET) scans from Magnetic Resonance Imaging (MRI). This pathology-aware method improves neurodegenerative disease diagnosis using MRI alone.
Area of Science:
- Medical Imaging
- Artificial Intelligence
- Neuroscience
Background:
- Positron Emission Tomography (PET) is crucial for diagnosing neurodegenerative diseases but is limited by cost and radiation.
- Magnetic Resonance Imaging (MRI) is safer and more accessible but less sensitive for detecting subtle pathological changes.
- Existing methods for generating synthetic PET from MRI often fail to capture crucial pathological details.
Purpose of the Study:
- To introduce PASTA, a novel framework for generating pathology-aware synthetic PET images from MRI.
- To enhance the diagnostic capabilities of MRI by improving its sensitivity to neurodegenerative pathologies.
- To overcome the limitations of current cross-modality medical image translation techniques.
Main Methods:
- Utilized conditional diffusion models for image translation with a focus on pathology awareness.
- Developed a dual-arm architecture and multi-modal condition integration for preserving structural and pathological details.
- Implemented a novel cycle exchange consistency and volumetric generation strategy for high-quality 3D PET synthesis.
Main Results:
- PASTA demonstrated superior performance over state-of-the-art methods in preserving both structural and pathological information.
- Synthesized PET scans achieved high qualitative and quantitative accuracy, exhibiting significant pathology awareness.
- For Alzheimer's disease diagnosis, synthesized PET scans improved performance by 4% compared to MRI alone, nearing actual PET performance.
Conclusions:
- PASTA effectively generates high-quality, pathology-aware synthetic PET images from MRI, addressing a critical gap in medical image translation.
- The framework holds significant potential for improving the diagnosis of neurodegenerative diseases by enhancing MRI's capabilities.
- PASTA offers a promising, cost-effective, and radiation-free alternative for functional brain imaging in clinical settings.
