Table 3.
Correlation results over various ranges of iron overload
| Mean (R2* 3D, R2* 2D) | (R2* 3D-R2* 2D) values | ||
|---|---|---|---|
| 0–200 s−1 | 0–400 s−1 | 0–600 s−1 | |
| Number of acquisitions | 35 | 47 | 54 |
| p value of Kolmogorow–Smirnow test* | 0.333 | 0.214 | 0.003 |
| Pearson’s correlation coefficient (95% CI) | 0.99 (0.97–0.99) | 0.99 (0.99–1.00) | 0.99 (0.97–0.99) |
| Lin’s concordance coefficient (95% CI) | 0.99 (0.97–0.99) | 0.99 (0.98–0.99) | 0.98 (0.97–0.99) |
| Slope (95% CI) | 0.98 (0.92–1.03) | 1.05 (1.01–1.09) | 0.94 (0.89–0.98) |
| Intercept (95% CI) | 2.69 (− 3.17–8.55) | − 1.86 (− 8.28–4.56) | 11.34 (0.50–22.17) |
Very strong correlation is noted across all levels of iron overload, however, for R2* values less than 400 s−1, excellent agreement with slope and intercept close to 1.0 and 0, respectively, demonstrate excellent reproducibility between 2 and 3D R2* mapping methods. Results of the analysis of the Kolmogorow–Smirnow test of normality, the Pearson’s correlation coefficient, Lin’ concordance coefficient and the slope and interception of the linear regression for three subgroups with R2* measurements are at 0–200 s−1, 0–400 s−1, and 0–600 s−1, respectively. R2* measurements of both the retrospective clinical and the prospective research studies are included in this analysis
CI confidence interval
*A p value less than 0.05 indicates the data differ significantly from the normal distribution