Deep Learning MRI Reconstruction for Fast Dynamic Image Sequences

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Dynamic MRI imaging is a slow process due to the need for acceleration methods to reconstruct images from incomplete data, necessitating more efficient and rapid image decoding and reconstruction techniques.

Innovation Solution

A method involving neural networks to process preliminary spatial weighting functions, extracting artifact-free final spatial weighting functions through training, and multiplying them with temporal basis functions to generate image sequences, utilizing auxiliary data and a specific neural network architecture for efficient image reconstruction.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If conventional acceleration methods are used to reconstruct images from incomplete imaging data, then image reconstruction can be achieved, but the process is slow and time-consuming

Engineering Contradiction:

Improveimage reconstruction speedVSAvoidreconstruction time

Core Design Contradiction:

ProductivityVSLoss of time

Solution Approach 1:

The patent extracts temporal basis functions from auxiliary data before the actual image reconstruction process. These pre-extracted basis functions are then used to accelerate the reconstruction of dynamic MRI image sequences, eliminating the need for time-consuming iterative reconstruction methods during the actual imaging process.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The patent replaces conventional iterative mathematical reconstruction algorithms with a neural network-based approach. The neural network is trained to directly map incomplete imaging data to reconstructed images, substituting the mechanical iterative computation process with a learned mapping that achieves faster reconstruction speeds.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

2Productivity

If acceleration methods are applied to reconstruct dynamic MRI images, then imaging speed improves, but image quality and accuracy deteriorate due to artifacts

Engineering Contradiction:

Improvedecoding speedVSAvoidimage reconstruction accuracy

Core Design Contradiction:

ProductivityVSManufacturing precision

Solution Approach 1:

The patent introduces temporal basis functions as an intermediary representation between the incomplete imaging data and the final reconstructed images. These basis functions, extracted from auxiliary data, serve as a bridge that preserves temporal coherence and reduces artifacts while enabling fast reconstruction through the neural network.

Inventive Principle:

Principle #24Intermediary (Mediator)

Solution Approach 2:

The patent combines multiple components including preliminary spatial weighting functions, temporal basis functions, and neural network processing to create a composite reconstruction approach. This composite method integrates the strengths of each component to achieve both high speed and high accuracy in dynamic MRI reconstruction.

Inventive Principle:

Principle #40Composite materials

Data Source

PatentEP4028787B1Systems and methods of deep learning for large-scale dynamic magnetic resonance image reconstruction

Publication Date: 2025.10.29 CEDARS SINAI MEDICAL CENT

Figure 1
Figure 2
Figure 3A~3B

AI summary

A method for performing magnetic resonance imaging on a subject comprises obtaining undersampled imaging data, extracting one or more temporal basis functions from the imaging data, extracting one or more preliminary spatial weighting functions from the imaging data, inputting the one or more preliminary spatial weighting functions into a neural network to produce one or more final spatial weighting functions, and multiplying the one or more final spatial weighting functions by the one or more temporal basis functions to generate an image sequence. Each of the temporal basis functions corresponds to at least one time-varying dimension of the subject. Each of the preliminary spatial weighting functions corresponds to a spatially-varying dimension of the subject. Each of the final spatial weighting functions is an artifact-free estimation of the one of the one or more preliminary spatial weighting functions.