Heart and liver T2* assessment for iron overload using different software programs

To assess the level of agreement and interchangeability among different software programs for calculation of T2* values for iron overload. T2* images were analysed in 60 patients with thalassaemia major using the truncation method in three software programs. Levels of agreement were assessed using Pearson correlation and Bland-Altman plots. Categorical classification for levels of iron concentration by each software program was also compared. For the heart, all correlation coefficients were significant among the software programs (P < 0.001 for all coefficients). The mean differences and 95% limits of agreement were 0.2 (-4.73 to 5.0); 0.1 (-4.0 to 3.9); and -0.1 (-4.3 to 4.8). For the liver all correlations were also significant with P < 0.001. Bland-Altman plots showed differences of -0.02 (-0.7 to 0.6); 0.01 (-0.4 to 0.4); and -0.02 (-0.6 to 0.6). There were no significant differences in clinical classification among the software programs. All tools used in this study provided very good agreement among heart and liver T2* values. The results indicate that interpretation of T2* data is interchangeable with any of the software programs tested. Magnetic resonance imaging in iron overload assessment has become an essential tool. Post processing options to establish T2* values have not been compared. No differences were found on T2* of the liver or heart using 3 different techniques. Availability of these methods should allow more widespread interpretation of iron overload by MRI.