PhaseGen: A Diffusion-Based Approach for Complex-Valued MRI Data Generation

2025-04-10

AI Summary

Existing MRI AI methods predominantly rely on magnitude images, neglecting diagnostically valuable phase information and thus failing to model the complex-valued k-space domain. To address this gap, we propose PhaseGen—the first conditional diffusion generative model explicitly designed for the complex-valued k-space domain in MRI. PhaseGen synthesizes high-fidelity phase data end-to-end, conditioned solely on clinically acquired magnitude images, thereby reconstructing missing complex-valued k-space entries. Our approach introduces controllable complex-domain generation in k-space and pioneers the integration of phase information into AI-based medical image pretraining, enabling pure k-space domain tasks such as skull-stripping. Built upon the DDPM framework, it employs complex-valued parameterization and k-space conditioning, ensuring compatibility with standard pipelines (e.g., FastMRI). Experiments demonstrate that skull-stripping accuracy improves from 41.1% to 80.1%, and performance gains are substantial even with limited measured phase data. Code is publicly available.

Technology Category

• Machine Learning: Deep Generative Models & Autoencoders
• Computer Vision: Diffusion Models for Vision
• Natural Language Processing: Generation

Application Category

• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Social Networks and Social Media: Generative AI / large language models and their impact on social systems
• Economics, Online Markets and Human Computation: Trust and reliance of crowd workers and data experts on GenAI

Abstract

Magnetic resonance imaging (MRI) raw data, or k-Space data, is complex-valued, containing both magnitude and phase information. However, clinical and existing Artificial Intelligence (AI)-based methods focus only on magnitude images, discarding the phase data despite its potential for downstream tasks, such as tumor segmentation and classification. In this work, we introduce $ extit{PhaseGen}$, a novel complex-valued diffusion model for generating synthetic MRI raw data conditioned on magnitude images, commonly used in clinical practice. This enables the creation of artificial complex-valued raw data, allowing pretraining for models that require k-Space information. We evaluate PhaseGen on two tasks: skull-stripping directly in k-Space and MRI reconstruction using the publicly available FastMRI dataset. Our results show that training with synthetic phase data significantly improves generalization for skull-stripping on real-world data, with an increased segmentation accuracy from $41.1%$ to $80.1%$, and enhances MRI reconstruction when combined with limited real-world data. This work presents a step forward in utilizing generative AI to bridge the gap between magnitude-based datasets and the complex-valued nature of MRI raw data. This approach allows researchers to leverage the vast amount of avaliable image domain data in combination with the information-rich k-Space data for more accurate and efficient diagnostic tasks. We make our code publicly $href{https://github.com/TIO-IKIM/PhaseGen}{ ext{available here}}$.

Problem

• Generates synthetic complex-valued MRI data from magnitude images
• Improves skull-stripping accuracy in k-Space using synthetic phase data
• Enhances MRI reconstruction by combining synthetic and limited real data

Innovation

• Diffusion model for complex-valued MRI data
• Generates synthetic MRI raw data
• Improves segmentation and reconstruction accuracy