The combination of multi-contrast quantitative MR imaging techniques may have great potential to improve the accuracy of tumor differentiation and grading compared to using one single imaging method. Previous work [1] reported that joint usage of ADC and T2 map shows significantly higher accuracy in tumor differentiation than using ADC or T2 alone, which could provide almost perfect separation of peripheral-zone prostate cancer and normal peripheral zone. In this study, we propose a fast method to acquire ADC and T2 maps simultaneously based on a SE-EPI sequence, which may facilitate the future application of a combined ADC and T2 evaluation method. To simultaneously model ADC and T2, the dataset should contain multiple b-values and multiple TE times. An integrated dataset is collected from 4 acquisitions using a prototype SE-EPI-based diffusion sequence, with different TE times from 40 to 100 ms, and each acquisition contains different b-value ranges, from 0 to 1000 s/mm² (Table 1). Other sequence parameters are: TR = 4500 ms, FOV 171 x 214 mm², slice thickness 3 mm, acquisition matrix 72 x 90. For calculation of ADC and T2, we used a two-step linear-least-square fitting method. First, T2 is calculated by fitting multiple TE data with same b-value using equation: S(b, TEn) = S(b, 0) · exp(- TEn / T2) , where S(b, TEn) is the signal at the nth TE with b-value = b, S(b,0) and T2 are the fitted variables. Second, the diffusion data with multiple TE times are extrapolated to TE = 58 ms using the T2 map at corresponding b-value, and are then averaged to increase signal-to-noise ratio. Afterwards, the ADC is calculated by fitting data with same TE (58 ms), using equation S(b, TE) = S(0, TE) · exp(-b · ADC).Eight prostate diffusion datasets (2 normal, 3 benign prostatic hyperplasia BPH, 3 prostate cancer PCa) were collected using a 3T system (Magnetom Skyra, Siemens Healthcare, Erlangen, Germany). The algorithms are implemented using MATLAB. Three T2 maps were calculated from SE-EPI sequence at b-value 0, 100 and 200 s/mm² as well as ADC map, and ROI analysis was performed (Figure 1 & 2). The results showed that both the T2 and ADC values of the BPH group were much higher than those of PCa and normal group (Table 2). For central zone, PCa group showed lower ADC than that from normal group, but with similar T2 values. In addition, we found that the T2 maps generated by the SE-EPI-based method were comparable to those from a conventional TSE-based method reported in previous studies [2] (Table 2), and for the SE-EPI-based method, the T2 value decreases slightly from b 0 to 200 s/mm² in normal and BPH case. To test the possibility for further acquisition acceleration, we also generated T2 maps at b 200 s/mm² with only 2 TE diffusion-weighted data and compared it to those from full 4 TE data. The result showed no significant difference between two T2 maps from paired t-test ( P = 0.66 ).

The proposed simultaneous ADC and T2 mapping based on the SE-EPI sequence was shown to be technically feasible in prostate within a feasible clinical acquisition time (total time around 3 minutes). The experiment showed that the proposed method generated T2 map similar to using a conventional TSE method, but with much shorter acquisition time. Between different groups of patients, BPH and PCa groups showed huge difference of both mean ADC and T2 values, indicating that joint evaluation of ADC and T2 may improve the lesion differentiation in difficult cases[1]. We also found that the mean T2 values estimated at b = 0 were higher than those estimated at b = 100 and 200 s/mm². We think this may be due to the influence of perfusion components, and needs to be considered in the future. A further study with a larger population will be conducted to consolidate the result.

Although the SE-EPI-based T2 mapping method is suboptimal compared to a conventional TSE-based method, and suffers from lower spatial resolution, contamination of T2* signal, EPI-based distortion, etc., it is highly efficient and generates consistent results compared to the conventional method. Thus, this fast T2 mapping method has an advantage in multi-contrast quantitative studies in combination with diffusion and other quantitative methods. In addition, the proposed method can also be easily combined with other diffusion models, such as DKI [3] and IVIM [4]. Our study also showed that only two TE data are sufficient to generate very stable T2 maps, thus the additional time for T2 mapping can be limited.