Introduction to a novel T2 relaxation analysis method SAME-ECOS: Spectrum Analysis for Multiple Exponentials via Experimental Condition Oriented Simulation

TL;DR

SAME-ECOS is a new machine learning-based method for analyzing multi-exponential T2 relaxation data, providing faster and more reliable spectra decomposition tailored to specific MR experimental conditions.

Contribution

It introduces a novel spectrum analysis method combining information theory and neural networks, improving T2 spectra estimation over traditional algorithms.

Findings

More reliable T2 spectra compared to NNLS

Faster analysis suitable for clinical use

Effective across different MR experimental conditions

Abstract

We propose a novel T2 relaxation data analysis method which we have named spectrum analysis for multiple exponentials via experimental condition oriented simulation (SAME-ECOS). SAME-ECOS, which was developed based on a combination of information theory and machine learning neural network algorithms, is tailored for different MR experimental conditions, decomposing multi-exponential decay data into T2 spectra, which had been considered an ill-posed problem using conventional fitting algorithms, including the commonly used non-negative least squares (NNLS) method. Our results demonstrated that, compared with NNLS, the simulation-derived SAME-ECOS model yields much more reliable T2 spectra in a dramatically shorter time, increasing the feasibility of multi-component T2 decay analysis in clinical settings.