Introduction

Diffusion-weighted imaging (DWI) with a high b value of around 1000 sec/mm² is commonly acquired to assess tumor cellularity and fluid viscosity in 3T MR examinations of the liver. Use of a higher b value is expected to improve this assessment but can degrade the image quality and interpretability with decreased signal-to-noise ratio (SNR) of the hepatic parenchyma. Recently, a state-of-the-art 3T MR scanner (Vantage Galan 3T/ZGO; Canon, Tochigi, Japan) with the maximal gradient magnetic field of 100 mT/m has been clinically introduced. This high gradient magnetic field allows the selection of an ultrahigh b value (e.g., b = 3000) without the need to increase echo time (TE) on liver DWI with preserving the SNR. The purpose of the present study was to assess impacts of different b and TE values on quality of liver DWI and to prove feasibility of the ultrahigh-b-value DWI using this MR scanner.

Methods

We originally generated a phantom consisting of seven tubes of 3-cm diameter containing gelatin/distilled water with various weight percentages (0, 10, 20, 30, 40, 45, and 50%), one at the center and six symmetrically placed around it, in a polyethylene container of 20-cm diameter filled with superabsorbent polymer (Figure 1). With this MR scanner, we scanned this phantom with various T2 (31.5-868.6 msec) and apparent diffusion coefficient (ADC) values (0.45-2.01 x 10^-3 mm²/sec) using body (Atlas SPEEDER body) and spine coils (Atlas SPEEDER spine) and acquired DWI using a spin-echo echo planner imaging sequence with the following scan parameters: repetition time (TR), 7000 msec; flip angle, 90/180 degrees; field of view, 33 x 36 cm; matrix size, 160 x 160; b value, 1000 and 3000 sec/mm²; TE, 44 (b = 1000), 53 (b = 3000), and 70 msec (b = 1000 and 3000). We placed regions of interest (ROIs) in the cross-section image of each tube phantom and background to measure the SNR and ADC values. We also scanned 10 volunteers (5 men and 5 woman; age, 69.1±19.0 years) and 11 patients who had hepatic masses (5 men and 6 women; age, 72.8±9.7 years) including hepatocellular carcinoma in one patient; metastatic tumor, one; focal nodular hyperplasia, one; hemangioma, two; complicated cyst, one; and simple cyst, five. We placed four ROIs in the both hepatic lobes in each volunteer and one ROI in a hepatic mass in each patient. We calculated SNR and ADC values at different b and TE values. We used paired t test to compare the SNR and ADC values of the hepatic parenchyma and masses between different TE values (TE 70 vs. TE 44 or 53) at the same b value.

Results

In the phantom study (Figure 1), the SNR at TE 44 was higher than that at TE 70 at b value of 1000. The SNR at TE 53 was higher than that at TE 70 at b value of 3000. The ADC was comparable between different TE values at the both b values. The mean SNR of the hepatic parenchyma in volunteers was significantly higher at TE 44 (94.7±15.8) than at TE 70 (35.9±7.1) at b value of 1000; and at TE 53 (18.9±4.4) than at TE 70 (12.4±3.2) at b value of 3000 (p < 0.05 for both) (Figure 2). The mean ADC was comparable between different TE values (p > 0.05 for both) (Figure 3). The mean SNR of the hepatic masses was also significantly higher at TE 44 (122.1±61.7) than at TE 70 (86.5±52.0) at b value of 1000; and at TE 53 (25.7±17.4) than at TE 70 (19.4±11.8) at b value of 3000 (p < 0.05 for both) (Figure 2). The mean ADC was comparable between different TE values (p > 0.05 for both) (Figure 3).

Discussion

Use of short TE values offers higher SNR on DWI than at TE 70, commonly used for liver DWI, without affecting ADC values; actually, can reveal the hepatic parenchyma even at ultrahigh b value of 3000 (Figure 4) and more accurately differentiate malignant (e.g., metastatic tumor) from benign hepatic masses (e.g., hemangioma) at the b value (Figure 5).

Conclusion

Use of short TE values can achieve sufficient SNR on liver DWI even at ultrahigh b value of 3000 with this MR scanner.

Acknowledgements

No acknowledgement found.

References

No reference found.