Nadeen K. Abu Ata

Background Magnetic resonance cholangiopancreatography (MRCP) is used to diagnose and monitor primary sclerosing cholangitis (PSC). Objective To assess inter-reader agreement for the diagnosis of PSC/autoimmune sclerosing cholangitis (ASC) and for individual MRCP features of cholangiopathy in a pediatric sample. Materials and methods This was a retrospective, IRB-approved study that included MRCP examinations from patients κ) statistics and intra-class correlation coefficients (ICC) with 95% confidence intervals (CI) were used to measure inter-reader agreement. Results Seventy-five patients were included (median age=16.8 [IQR 13.8-18.7] years; 48 boys); 22.7% (17/75) had PSC, 22.7% (17/75) had ASC, and 54.7% (41/75) had other diagnoses. Among observers, agreement was only slight for presence of cholangiopathy (κ=0.15 [95% CI 0.07 to 0.23]) and presence of PSC/ASC (κ=0.13 [0.06 to 0.21]). Agreement was poor for categorical intrahepatic stricture number (κ=-0.002 [(-0.16 to 0.15]) and stricture extent (κ=-0.06 [-0.09 to-0.02]). Agreement was slight for presence of intrahepatic stricturing disease (κ=0.08 [0.04 to 0.12]). Most other findings had fair agreement between readers (including intrahepatic focal dilations, intrahepatic and extrahepatic diverticula, diffuse extrahepatic dilation without stricture, bile duct mural thickening, and biliary obstruction [κ=0.22 to 0.34]). There was moderate agreement for categorical extrahepatic stricture length (κ=0.46 [-0.11 to 1]) and presence of extrahepatic biliary dilation (κ=0.53 [0.40 to 0.65]). There was excellent agreement for extrahepatic bile duct maximum diameter (ICC=0.89 [0.85 to 0.92]). Conclusion Inter-reader agreement for interpreting MRCP in children and young adults is slight to fair for the diagnosis of PSC/ASC and for most findings of cholangiopathy. Graphical Abstract

Objectives Implementation of diffusion-weighted imaging (DWI) for abdominal imaging in children has challenges due to motion artifacts exacerbated by long acquisition times. We aimed to compare acquisition time and image quality between conventional DWI and multi-band (MB) DWI of the liver in children and young adults. Methods Clinical MRI exams from May 2023 to January 2024 were reviewed, including four DWI sequences: respiratory-triggered (RTr, clinical standard), free-breathing (FB), MB-DWI with shift factor 1 (MBsf1), and MB-DWI with shift factor 2 (MBsf2). Acquisition times were recorded, and signal intensity and apparent signal-to-noise ratio (aSNR) were calculated for the liver and spleen. Six blinded pediatric radiologists independently assessed image quality, artifacts, and lesion visualization on a 5-point Likert scale and identified their preferred sequence. Statistical comparisons were made using Kruskal-Wallis and ANOVA tests. Results Median acquisition times were significantly reduced with MB-DWI (43 s for MBsf1/MBsf2) compared to FB (84 s) and RTr (240 s). Image quality and artifact scores were highest for RTr and FB sequences (p < 0.0001). Mean image quality scores were 3.7 (RTr, FB), 3.4 (MBsf1), and 3.5 (MBsf2), while artifact scores followed a similar trend (higher score = fewer artifacts). Lesion visualization scores were comparable across sequences (p = 0.11), and reviewers expressed no preference in 47% of cases. Apparent diffusion coefficient (ADC) values were consistent across all sequences (p > 0.05). Conclusion MB-DWI significantly reduces acquisition time while maintaining acceptable image quality and lesion visualization, making it a valuable option for pediatric abdominal MRI.