Training deep learning based dynamic MR image reconstruction using synthetic fractals

2026-03-31

AI Summary

This study addresses the challenge of limited access to real clinical cardiac MRI data due to privacy and acquisition constraints by proposing a fully synthetic-data-driven deep learning reconstruction approach. The method leverages quaternion-based Julia fractals to generate 2D+time dynamic images, which are combined with multi-coil MRI simulation and radial undersampling to produce paired k-space data for training a 3D U-Net model to suppress artifacts. For the first time, it is demonstrated that a fractal-based deep learning (F-DL) model trained exclusively on synthetic fractal data achieves image quality statistically indistinguishable from that of models trained on real patient data (subjective rating p=0.9) and yields clinically relevant metrics—such as ventricular volumes and ejection fraction—without significant bias. Moreover, the F-DL model significantly outperforms both compressed sensing and low-rank deep prior methods (p<0.001), establishing a novel paradigm for open, scalable MRI reconstruction without reliance on real patient data.

Technology Category

• Machine Learning: Multimodal Learning
• Natural Language Processing: Code Generation / Program Synthesis from Natural Language
• Computer Vision: Diffusion Models for Vision

Application Category

• User Modeling, Personalization and Recommendation: Federated recommendation systems and personalization
• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Purpose: To investigate whether synthetically generated fractal data can be used to train deep learning (DL) models for dynamic MRI reconstruction, thereby avoiding the privacy, licensing, and availability limitations associated with cardiac MR training datasets. Methods: A training dataset was generated using quaternion Julia fractals to produce 2D+time images. Multi-coil MRI acquisition was simulated to generate paired fully sampled and radially undersampled k-space data. A 3D UNet deep artefact suppression model was trained using these fractal data (F-DL) and compared with an identical model trained on cardiac MRI data (CMR-DL). Both models were evaluated on prospectively acquired radial real-time cardiac MRI from 10 patients. Reconstructions were compared against compressed sensing(CS) and low-rank deep image prior (LR-DIP). All reconstrctuions were ranked for image quality, while ventricular volumes and ejection fraction were compared with reference breath-hold cine MRI. Results: There was no significant difference in qualitative ranking between F-DL and CMR-DL (p=0.9), while both outperformed CS and LR-DIP (p<0.001). Ventricular volumes and function derived from F-DL were similar to CMR-DL, showing no significant bias and accptable limits of agreement compared to reference cine imaging. However, LR-DIP had a signifcant bias (p=0.016) and wider lmits of agreement. Conclusion: DL models trained using synthetic fractal data can reconstruct real-time cardiac MRI with image quality and clinical measurements comparable to models trained on true cardiac MRI data. Fractal training data provide an open, scalable alternative to clinical datasets and may enable development of more generalisable DL reconstruction models for dynamic MRI.

Problem

• dynamic MRI reconstruction
• deep learning
• synthetic data
• fractal
• cardiac MRI

Innovation

• synthetic fractals
• dynamic MRI reconstruction
• deep learning
• data-free training
• quaternion Julia fractals