Synopsis

Using combination of compressed sensing and parallel imaging for arterial phase image acquisition in gadoxetic acid enhanced liver MR

Purpose: To evaluate whether combination of compressed sensing and two-dimensional parallel imaging can reduce the scan time of triple arterial phase in gadoxetic acid-enhanced MR compared to the use of parallel imaging alone

Materials and Methods: 253 consecutive patients who underwent gadoxetic acid enhanced liver MR using a 3T MR scanner (Ingenia CX; Philips Healthcare) were included in this study. Triple arterial phase (TAP) images were acquired in all patients with two different method: combination of compressed sensing (CS) and two-dimensional parallel imaging (PI) with the scan time of 15 seconds (n=127); and the use of two-dimensional PI alone with scan time of 18 seconds (n=126). Arterial phase timing was evaluated, and the degree of motion related artifact as well as image quality were assessed by using four-point scale. Scores for motion related artifact and image quality were compared between two methods using Mann-Whitney U test. The rate of successful late arterial phase acquisition without significant motion (score ≤ 2) was also calculated and compared between two methods using the chi-square test.

Results: Motion related artifact was not significantly different between two acquisition methods (3.2 for combination of CS and PI vs. 3.3 for PI alone, P=0.136). Image quality of arterial phase image obtained using combination of CS and PI (3.2) was not significantly different from PI alone (3.3) (P=0.168). TAP using combination of CS and PI captured late arterial phase without significant motion in 87.4% (111/127), and triple arterial phase using PI alone in 92.1% (116/126): this difference was not statistically significant (P=0.22).

Conclusion: Combination of CS and PI can reduce the scan time for TAP of gadoxetic acid enhanced liver MR providing comparable imaging quality to the use of PI alone.

Acknowledgements

References