Synthesizing MRI Contrast from Three Standard Sequences
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current MRI technologies require multiple sequences and longer scan times to produce high-quality images, which increases healthcare costs and reduces patient throughput.
Innovation Solution
A method that synthesizes MR images of varying contrast using numerical input for varying scan parameters, allowing for the rapid generation of quantitative MR parametric maps from just three standard artefact-free MR sequences, including historical data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple MRI sequences are acquired to produce high-quality images, then image quality is improved, but scan time increases
Solution Approach 1:
The patent applies preliminary action by acquiring only three standard MRI sequences with specific scan parameters (TR/TE combinations) that are pre-determined to be sufficient for generating quantitative parametric maps. This preliminary acquisition strategy eliminates the need for multiple additional sequences, thereby reducing scan time while maintaining image quality through the use of optimized parameter combinations rather than exhaustive sequencing.
Solution Approach 2:
The patent utilizes parameter changes by varying TR and TE values across the three acquired sequences to generate different tissue contrast characteristics. By systematically changing these scan parameters, the method extracts quantitative relaxation information (T1, T2, proton density) without requiring multiple separate sequence types, thus reducing scan time while preserving diagnostic quality through parameter-based differentiation.
2Measurement precision
If multiple MRI sequences are acquired to obtain comprehensive tissue information, then diagnostic accuracy is improved, but patient throughput decreases
Solution Approach 1:
The patent applies universality by designing a multi-functional approach where three standard MRI sequences serve multiple purposes: they simultaneously provide T1-weighted, T2-weighted, and proton density information, and can be used to generate various contrast-weighted images through post-processing. This multi-functionality eliminates the need for separate dedicated sequences for each tissue characteristic, thereby maintaining diagnostic accuracy while increasing patient throughput by reducing total scan time.
Solution Approach 2:
The patent employs copying by creating synthesized MR images with different contrast weights (T1-weighted, T2-weighted, PD-weighted) from the three acquired sequences using quantitative parametric maps. These synthesized images replicate the information that would otherwise require separate acquisition sequences, thereby maintaining diagnostic comprehensiveness while reducing scan time and increasing patient throughput.
3Measurement precision
If conventional MRI sequences are used to ensure high image quality, then image contrast is improved, but scan time increases
Solution Approach 1:
The patent applies parameter changes by acquiring three sequences with specific TR/TE parameter combinations that optimize tissue contrast characteristics. By carefully selecting these parameters (e.g., short TR/TE for T1, long TR/TE for T2), the method achieves high image contrast without requiring multiple separate sequence types, thus reducing scan time while maintaining diagnostic quality through parameter-optimized contrast enhancement.
Data Source
AI summary
The present invention relates to a method for producing MRI images with the following features: Magnetic resonance images are acquired using an MRI system. These images can be used to determine the relaxation times of protons and can be displayed as parametric maps (110, 111, 112). Various other contrasts can be synthesized by combining T1, T2 and PD values and varying the scan parameters using Bloch equations or modifications (104). The images thus synthesized will be displayed in the appropriate display. The entire process can be implemented even as a client-side software programme or even as a web based application.