Methods for solving the neck two-point T₂-DIXON fat-water separation misalignment

Objective: To explore the methods for solving the neck two-point T₂-DIXON fat-water separation misalignment. Methods: During August 2015 to July 2016, 140 patients with fat-water separation on the nasopharynx and cervical axial T₂-DIXON images were prospectively recruited from Sichuan Cancer Hospital. There was no metal implant in the fat-water separation misalignment area. The patients were divided into 7 groups by random number table method with 20 patients in each group for the axial T₂-DIXON scan: Group A adopted plan 1 (increasing localized shim box on the fat-water separation misalignment area); Group B adopted plan 2 (increasing the times of repeated acquisition to 2 of the T₂-DIXON sequence); Group C adopted plan 3 (placing the shimming assist device on the inspection area); Group D adopted plan 1 + plan 2; Group E adopted plan 1 + plan 3; Group F adopted plan 2 + plan 3; Group G adopted plan 1 + plan 2 + plan 3. The images quality of two scans were graded and compared. We compared the signal-to-noise ratio (SNR) of T₂-DIXON image on the same muscle tissue between the two scans of each group. The difference of SNR on the two scan images was compared with the paired t test, the difference of SNR among seven groups was conducted with independent sample t test, and the comparison of image quality classification was conducted by rank sum test. Results: The image quality of all the seven groups was improved to some different degrees. The cases with image quality reaching level 3 were 12, 15, 15, 16, 17, 18 and 18 in A to G groups, respectively. It was better to use the combination of two or more methods to improve the quality of the image than to use a single method. There were no statistically significant differences in SNR between two scans in A, C and E group (all P > 0.05). There were statistically significant differences in SNR between two scans in groups of B, D, F and G (all P < 0.05). There were statistically significant differences among the 7 groups, with the best quality in G group (P < 0.05). There was no significant difference in the first SNR among the 7 groups (P > 0.05); there were significant differences in the second scans among the 7 groups (P < 0.05). Conclusions: This study suggests that placing localized shim box, increasing the times of repeated acquisition, and use of shimming assist device in MRI correct the fat-water separation misalignment, help to provide images with high quality. The combination of the above method was better than using the single method. The SNR can be improved when increasing the times of repeated acquisition. (authors)