Recent advances in MRI in the preoperative assessment of anorectal malformations

AbstractObjectiveTo prospectively evaluate the diagnostic accuracy of triplanar 2D magnetic resonance (MR) images versus the post processing reconstructed MR images generated from a single 3D VISTA sequence in children with anorectal malformations compared with the operative findings as the Golden standard.Materials and methodsThis study had institutional review board approval, and informed consent was obtained from all participants. There were 20 patients with anorectal anomalies age range (2months–3.5years), referred from the pediatric surgery unit for preoperative MRI. The 2D sequences included Coronal TSE T2 WI parallel to the plane of the anal canal as seen on the sagittal images, and Axial TSE T2 WI oblique, perpendicular to the coronal plane. Sagittal T2 weighted 3D-sequence; volumetric isotropic turbo spin echo acquisition (VISTA) was taken parallel to the mid-sagittal plane.ResultsComparing the two MR imagining protocols the acquisition time of the sagittal 3D sequence (6min 42s) was shorter than the total acquisition time of the 2D sequences (10min 12s) without jeopardizing the image quality or accuracy of the diagnostic information. MRI depicted the fistula in 94% (16 out of 17) of patients who were clinically proven fistulous, which is a privilege to detect associated anomalies on the same examination. The frequency of associated anomalies was 100% in high anorectal malformations (ARMs), 64% in intermediate and only 20% in low ARMs. From these associated anomalies vertebral anomalies were seen in 50% of patients. Concomitant cord anomalies were encountered in 60% of those with vertebral anomalies the most common of which was tethered cord observed in 35% of the patients.ConclusionMRI is a valuable diagnostic tool to diagnose type and level of ARM in addition to the privilege of associated anomalies detection on the same examination. The 3D sequence acquisition with multiplanar reconstruction has the advantage of time saving without compromise on image quality or the diagnostic accuracy compared to 2D images