Clinical usefulness of free-breathing navigator-triggered 3D MRCP in non-cooperative patients: Comparison with conventional breath-hold 2D MRCP

Abstract

To assess the clinical usefulness of free-breathing 3D MRCP in non-cooperative patients compared conventional breath-hold 2D MRCP. We performed FB navigator-triggered 3D MRCP using prospective acquisition correction and BH 2D MRCP in 48 consecutive, non-cooperative patients among 772 patients. Thirteen patients had malignant obstruction. Two radiologists independently graded the likelihood of a malignant obstruction, the overall image quality, and the visibility of ten, individual anatomic segments of both the biliary and pancreatic duct in each sequence. The area under the ROC curve and the repeated measures analyses of variance with multiple comparisons were used for the comparison. The κ statistics were used for interobserver agreement. The diagnostic performance for detecting malignancy was significantly higher on FB MRCP (A(z)=0.962) than on either BH SS-RARE (A(z)=0.820, P<0.0185) or MS-HASTE MRCP (A(z)=0.816, P<0.0067). Interobserver agreement was excellent for FB MRCP (κ=0.889) and fair for both BH SS-RARE (κ=0.578) and MS-HASTE MRCP (κ=0.49). FB MRCP had a significantly higher technical quality than BH MRCP (P<0.001). FB MRCP was seen to have statistically better visibility of peripheral IHD, right main IHD, CHD, cystic duct, and CBD than BH MRCP (P<0.001). FB MRCP and BH SS-RARE MRCP had statistically better visibility of both the left main IHD and pancreatic duct than did BH MS-HASTE MRCP (P<0.001). FB 3D MRCP is useful for non-cooperative patients in whom conventional BH 2D methods cannot be used successfully.