Dark blood versus bright blood T2 acquisition in cardiovascular magnetic resonance (CMR) for thalassaemia major (TM) patients: evaluation of feasibility, reproducibility and image quality.

To compare the effectiveness of dark blood (DB) versus bright blood (BB) sequences. To assess the intra and inter-observer variability and inter-study reproducibility between BB versus DB. To evaluate image quality level in the two sequences.In a setting of 138 patients we performed CMR using cardiac gated Gradient-multiecho single breath-hold BB and DB sequences in the middle ventricular septum. Each acquisition was repeated during the same exam. Truncation method was used to account for background noise. Image quality (IQ) was assessed using a 5 point grading scale and image analysis was conducted by 2 experienced observers.Compared with the conventional BB acquisition, the coefficient of correlation and significance of the DB technique was superior for intra-observer reproducibility (p<0.001), inter-observer reproducibility (p<0.001) and inter-study reproducibility (p<0.001). The variability is also lower for DB sequences for T2* values <14 ms. Assessment of artifacts showed a superior score for DB versus BB scans (4 versus 3, p<0.001).Improvement in terms of inter observer and inter study variability using DB sequences was obtained. The greatest disparity between them was seen in inter-study reproducibility and higher IQ in DB was seen. Study demonstrates better performance of DB imaging compared to BB in presence of comparable effectiveness.