Magn Reson Med. 2025 Nov 2. doi: 10.1002/mrm.70078. Online ahead of print.
ABSTRACT
PURPOSE: The principles of signal-intensity-ratio (SIR) estimation were revisited to derive a set of generalizable equations to quantify liver iron concentration (LIC), extending the effective dynamic range of LIC-R2* estimation or detect when LIC- R2* estimation degenerates.
THEORY AND METHODS: Equations for LIC were derived for two techniques: (1) relaxometry of multi-echo gradient echo and (2) SIR of liver and muscle tissue in a single echo. MRI was completed at 1.5T and 3T in human subjects with a range of liver iron burdens. LIC was estimated by both techniques and compared to assess concordance across the clinical range and high LIC performance.
RESULTS: Regression of Log(SIR) by LIC was linear for both 1.5T and 3T data with similar slopes. Comparison between SIR-LIC and R2*-LIC estimates demonstrated linear concordance until the echo time-based saturation of R2* estimates were reached, at which point SIR-LIC estimates continued to rise without a concomitant increase in R2*-LIC. Lack of spatial intensity correction caused systematic bias and increased variance in LIC estimates compared to intensity-corrected images.
CONCLUSIONS: LIC-SIR measurements can potentially be used to estimate liver iron concentration in examinations where R2* acquisitions were not performed or where the dynamic range of a multi-echo protocol is exceeded. Using spatial intensity correction where possible improves SIR-LIC estimates.
PMID:41177944 | DOI:10.1002/mrm.70078