Multi-Parameter Molecular MRI Quantification using Physics-Informed Self-Supervised Learning

2024-11-10

Communications Physics

AI Summary

To address the computational inefficiency and clinical impracticality of biophysical model fitting in molecular MRI, this work proposes a differentiable numerical solver framework that integrates ordinary differential equation (ODE) modeling with automatic differentiation. Methodologically, we reformulate the ODE solver into a differentiable analytic stepping scheme, enabling gradient-based end-to-end parameter estimation; we further introduce the first single-scan self-supervised training strategy for molecular MRI, eliminating the need for paired ground-truth labels or synthetic data. Our approach unifies physics-informed neural networks with the train-by-fit paradigm, specifically applied to magnetization transfer (MT) and chemical exchange saturation transfer (CEST) modeling. Experiments demonstrate whole-brain multiparametric quantification at 18.3 ± 8.3 minutes per subject, with cross-subject inference completed in just 1.0 ± 0.2 seconds. Estimated parameters show strong agreement with both literature values and conventional fitting results, markedly enhancing clinical translatability.

Technology Category

• Cognitive Modeling & Cognitive Systems: Neural Spike Coding
• Machine Learning: Bio-inspired Learning
• Search and Optimization: Mixed Discrete/Continuous Search

Application Category

• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

Biophysical model fitting plays a key role in obtaining quantitative parameters from physiological signals and images. However, the model complexity for molecular magnetic resonance imaging (MRI) often translates into excessive computation time, which makes clinical use impractical. Here, we present a generic computational approach for solving the parameter extraction inverse problem posed by ordinary differential equation (ODE) modeling coupled with experimental measurement of the system dynamics. This is achieved by formulating a numerical ODE solver to function as a step-wise analytical one, thereby making it compatible with automatic differentiation-based optimization. This enables efficient gradient-based model fitting, and provides a new approach to parameter quantification based on self-supervised learning from a single data observation. The neural-network-based train-by-fit pipeline was used to quantify semisolid magnetization transfer (MT) and chemical exchange saturation transfer (CEST) amide proton exchange parameters in the human brain, in an in-vivo molecular MRI study (n = 4). The entire pipeline of the first whole brain quantification was completed in 18.3 $pm$ 8.3 minutes. Reusing the single-subject-trained network for inference in new subjects took 1.0 $pm$ 0.2 s, to provide results in agreement with literature values and scan-specific fit results.

Problem

• Accelerates molecular MRI parameter quantification via self-supervised learning
• Reduces computation time for ODE-based biophysical model fitting
• Enables clinical use of complex MRI parameter extraction

Innovation

• Physics-informed self-supervised learning for MRI quantification
• ODE solver with automatic differentiation optimization
• Neural-network-based train-by-fit pipeline for rapid inference