Cardiac R2* values are independent of the image analysis approach employed

Area of Science:

• Cardiovascular Magnetic Resonance (CMR) imaging
• Biomedical data analysis
• Quantitative MRI techniques

Background:

• Myocardial R2* quantification is crucial for assessing iron overload.
• Different signal decay models exist for R2* calculation.
• Understanding model-specific differences is essential for consistent clinical interpretation.

Purpose of the Study:

• To compare myocardial R2* values derived from two decay models: truncation and exponential+constant (Exp-C).
• To assess the impact of different fitting approaches (single ROI vs. pixelwise) on R2* measurements.
• To evaluate the generalizability of findings across single and multicenter cohorts.

Main Methods:

• Comparison of R2* values from truncation and Exp-C models using single-center black/bright blood sequences.
• Assessment of generalizability with a multicenter cohort of mixed sequences.
• Utilized CMRtools for truncated exponential estimates (single ROI) and a pixelwise approach for Exp-C estimates.

Main Results:

• No significant differences in R2* values were observed between black and bright blood sequences.
• Both fitting algorithms showed high agreement (R-squared > 0.997, CV 3-4%).
• A minor systematic bias (~3%) was noted with pixelwise Exp-C in severe iron deposition, which resolved using a single ROI Exp-C approach, indicating pixelwise mapping as the source of bias.

Conclusions:

• Cardiac R2* values are independent of the signal decay model used within clinically relevant ranges.
• Clinicians can confidently compare R2* results across centers employing different calculation methods.
• The choice of fitting approach (single ROI vs. pixelwise) can influence measurements in cases of severe iron overload.