Fast spin-echo (FSE) and gradient- and spin-echo (GRASE) in fast MRI of the pelvis

Comparisons between image quality and diagnostic performance of 2D- and breath-hold 3D magnetic resonance cholangiopancreatography at 3T

OBJECTIVES

To compare the image quality and diagnostic performance of 2D MRCP to those of breath-hold 3D MRCP using compressed sensing (CS-MRCP) and gradient and spin-echo (GRASE-MRCP) at 3T.

METHODS

From January to November 2018, patients who underwent pancreatobiliary MRI including 2D MRCP and two breath-hold 3D MRCP using CS and GRASE at 3T were included. Three radiologists independently evaluated image quality, motion artifact, and pancreatic cyst conspicuity. Diagnostic performance was assessed for bile duct anatomic variation, bile duct, and pancreatic diseases using a composite algorithm as reference standards. Pancreatic lesion detectability and conspicuity were evaluated using JAFROC and generalized estimating equation analysis.

RESULTS

One hundred patients (male = 50) were included. Bile duct anatomic variation, bile duct and pancreatic diseases were present in respectively 31, 15, and 79 patients. Breath-hold 3D MRCP provided better image quality than 2D MRCP (3.5 ± 0.6 in 2D MRCP; 4.0 ± 0.7 in GRASE-MRCP and 3.9 ± 0.8 in CS-MRCP, p < 0.001 for both). There was no difference in motion artifact between 2D and breath-hold 3D MRCP (p = 0.1). Breath-hold 3D CS-MRCP provided better pancreatic cyst conspicuity than 2D MRCP (2.7 [95% CI: 2.5-3.0] vs. 2.3 [95% CI: 2.1-2.5], p = 0.001). There were no significant differences between the diagnostic performance of the three sequences in the detection of bile duct anatomic variation or pancreatic lesions (p > 0.05).

CONCLUSION

Breath-hold 3D MRCP with GRASE or CS can provide better image quality than 2D MRCP in a comparable scan time.

KEY POINTS

• Breath-hold 3D MRCP using compressed sensing (CS) or gradient and spin-echo (GRASE) provided a better image quality with less image blurring than 2D MRCP. • There were no significant differences between 2D MRCP and breath-hold 3D MRCP in either motion artifact or the number of non-diagnostic exams. • There were no significant differences between 2D MRCP and either type of breath-hold 3D MRCP in the diagnosis of bile duct anatomic variation or detection of pancreatic lesions.

Comparison of two-dimensional gradient echo, turbo spin echo and two-dimensional turbo gradient spin echo sequences in MRI of the cervical spinal cord anatomy

The aim of this study was to assess the detectability and distinguishability of the cervical spinal cord, the anterior and posterior spinal roots and of the internal anatomy of the cord (distinction of grey and white matter). For this purpose 20 healthy volunteers were examined using a 1.5 T MR unit with 20 mT/m gradient strength and a dedicated circular polarized neck array coil. Three T2* weighted (w). 2D gradient echo sequences, two T2 w. 2D turbo spin echo (TSE) sequences and one T2 w. 2D turbo gradient spin echo (TGSE) sequence were compared. The multiecho 2D fast low angle shot (FLASH) sequence with magnetization transfer saturation pulse (me FLASH+MTS) yielded the best results for liquor/compact bone, liquor/spinal cord and grey/white matter contrast, as found with regions of interest (ROI) analysis. The single echo 2D FLASH sequence was significantly poorer than the two me FLASH+/-MTS sequences. Two-dimensional TGSE as well as 2D TSE with a 256 matrix and with a 512 matrix yielded the poorest results. In the visual analysis the contrast between liquor and compact bone, liquor and cord as well as liquor and roots was best with me FLASH+MTS, whereas grey/white matter distinction was best using me FLASH-MTS. In conclusion, we would therefore recommend the inclusion of an axial T2* w. multiecho 2D spoiled gradient echo sequence with magnetization transfer saturation pulse and gradient motion rephasing in a MR imaging protocol of the cervical spine.