T2* mapping can directly quantify the iron load of tissue, which is closely related to numerous diseases, such as thalassemia, hereditary hemochromatosis and sickle cell disease. However, blood signal would induce artifacts which lead to T2* estimation inaccurate. To address this issue, a novel black-blood T2* mapping technique utilizing Delay Alternating with Nutation for Tailored Excitation (DANTE) preparation module⁽¹⁾ followed by multi-echo gradient echo (GRE) readout (DANTE-GRE) was developed to acquire more accurate T2* maps.To eliminate the artifacts originating from flowing blood by effectively suppressing blood signal, consequently, to acquire accurately black-blood T2* map.

Figure 1 shows the sequence diagram of DANTE-GRE. The DANTE preparation module is followed by multi-echo GRE readouts to sample ten images with different contrasts in one TR.

To testify the accuracy of T2* map obtained by DANTE-GRE, both phantom and in vivo experiments were conducted on a 3T scanner (Siemens Tim Trio, Germany), and compared with those obtained by Motion-Sensitive Driven Equilibrium (MSDE) prepared multi-echo GRE readout (MSDE-GRE) ⁽²⁾ and conventional GRE.

In the phantom experiments, five SPIO solutions with different concentrations of 0.05 mmol/L, 0.08 mmol/L, 0.10 mmol/L, 0.15 mmol/L and 0.20 mmol/L, were mapped. Two groups of parameters for the DANTE module were conducted. The first one was to fix the pulse train length as 80 and vary the pulse flip angle from 10° to 20°. While the second one was to fix the pulse flip angle as 15° and vary the pulse train length from 25 to 150. For comparison, MSDE-GRE and conventional GRE were also conducted. The field of speed (FOS) for MSDE-GRE were set to 10 cm/s, 15 cm/s, 20 cm/s, 30 cm/s, 50 cm/s and 75 cm/s, respectively. The same parameters for the GRE readout were used for above experiments: TR’ = 300 ms and T_es = 49.5 ms. TEs ranged from 1.62 ms to 64.62 ms with gap 7 ms.

In the in vivo experiments, 12 healthy volunteers (5 males and 7 females) were recruited, and 7 of them were randomly chosen for neck scan and the others for lower limb scan. Parameters for the neck scan included: DANTE pulse flip angle of 15° and pulse train length of 80. GRE repetition time TR' of 300 ms and echo space T_es of 50.4 ms. TEs ranging from 1.62 ms to 46.62 ms with gap of 5 ms. Voxel size of 0.7×0.7×2 mm3, GRE flip angle of 30° and averages of 3. Parameters for the lower limb scan were: DANTE pulse flip angle of 15° and pulse train length of 100. GRE repetition time TR' of 300 ms and echo space T_es of 49.5 ms. TEs ranging from 1.4 ms to 46.4 ms with gap of 5 ms. Voxel size of 1.1×1.1×3 mm3, GRE flip angle of 40° and averages of 3. For MSDE-GRE and convertional GRE, the same parameters of GRE readout as those of DANTE-GRE were used. And the FOS of 37 cm/s was used for MSDE-GRE.

T2* maps acquired in all experiments were calculated by fitting ten images to a three-parameter model $$$S_{i}=A\cdot{exp(-TE_{i}/T_2^*)}+C$$$, where S_i represents the sampled signal intensity, A is the signal amplitude and C is used to fit the noise. Levenberg-Marquardt algorithm in MATLAB was utilized to solve the above function.Phantom experiments validate the superiority of DANTE-GRE over MSDE-GRE in acquiring less deteriorated maps (Figure 2a-c) because it provides higher SNR. Generally, quantitative analysis (Table 1) shows that T2* values calculated from DANTE-GRE and MSDE-GRE are smaller than GRE. However, DANTE-GRE calculated T2* values own smaller differences to GRE than MSDE-GRE. Moreover, standard deviations of DANTE-GRE are larger than GRE while smaller than MSDE-GRE in most cases. In the in vivo experiments, DANTE-GRE own better performance in suppressing signal from slow flowing blood than MSDE-GRE (Figure 3a). Images acquired by DANTE-GRE display much higher SNR than those of MSDE-GRE, even though lower than those of GRE (Figure 3a and 4a). Additionally, T2* maps acquired by DANTE-GRE from both neck and lower limb scans show more homogeneity than those by MSDE-GRE and GRE with less noise points around flow blood and muscle area.Accurate T2* maps are obtained by DANTE-GRE due to its high SNR and effective blood signal suppression. Both the phantom and in vivo experiments demonstrated that DANTE-GRE may outperform MSDE-GRE and conventional GRE for T2* mapping.