Efficient estimation of gadolinium‐based contrast agent concentration using transient‐state keyhole MR‐STAT

TL;DR

This paper introduces a faster MRI method to estimate contrast agent concentration using a new accelerated protocol, tested in phantoms and synthetic patient data.

Contribution

A novel accelerated MR-STAT protocol that enables efficient and accurate GBCA concentration estimation in clinically feasible scan times.

Findings

The accelerated MR-STAT protocol achieved accurate T1, T2, and GBCA concentration measurements with an undersampling factor of 4.

Estimated GBCA concentrations showed a strong linear relationship with reference values (slope 1.034, intercept 0.009).

Abstract

Fast quantitative MRI (qMRI) techniques have emerged to quantify multiple tissue parameters from a single acquisition in clinically feasible scan times. To integrate fast qMRI techniques into a clinical MRI protocol, one needs to consider imaging after injection of a contrast agent (e.g., dynamic contrast‐enhanced [DCE] imaging). However, fully repeating the multi‐parametric scan after the injection requires lengthy scan times, which are impractical for tracking the contrast agent concentration.

This work aimed to develop and validate an accelerated multi‐parametric method capable of estimating the concentration of Gadolinium‐based contrast agents (GBCAs) in a more time‐efficient way.

We designed an accelerated 2D Magnetic Resonance Spin TomogrAphy in Time‐domain (MR‐STAT) protocol and tested it on a 3T MRI scanner. This is achieved by combining transient‐state acquisitions with Cartesian keyhole‐based acceleration. Based on the reconstructed T
1 values, GBCA concentrations were quantified in phantoms containing water, manganese chloride, and five different concentrations of gadobutrol (Gadovist), ranging from 0.05 to 0.9 mM. Linear regression analysis between the estimated GBCA concentration and the reference values was performed to assess the accuracy. The GBCA concentration quantification method was then applied for synthetic patient data and a hybrid human/phantom study to demonstrate the feasibility of this method.

The accelerated MR‐STAT protocol, with an undersampling factor of 4, provided accurate T
1, T
2, and GBCA concentration measurements. Estimated GBCA concentrations were in a strong linear relationship with reference values, with a slope and intercept on simple linear regression analysis of 1.034 and 0.009, respectively.

We present a multi‐parametric approach for quantitatively assessing the concentration of gadolinium in vitro and in a hybrid in vivo/in vitro controlled setup using the accelerated MR‐STAT protocol. The concentration of gadobutrol in the range of 0.05–0.9 mM could be measured in a clinically applicable scan time using the proposed method, but needs further in vivo validation.

Figures