Inter-reader reliability of functional liver imaging score derived from gadoxetic acid-enhanced MRI: a meta-analysis

Purpose: This study aimed to systematically determine the inter-reader reliability of the functional liver imaging score (FLIS) and explore the factors affecting it. Methods: Original articles reporting the inter-reader reliability of FLIS derived from gadoxetic acid-enhanced magnetic resonance imaging (MRI) were systematically searched in the MEDLINE and EMBASE databases from January 2013 to June 2022. Data synthesis was performed to calculate the meta-analytic pooled estimates of the FLIS and its three subcategories, including enhancement quality score (EnQS), excretion quality score (ExQS), and portal vein sign quality score (PVsQS) using the DerSimonian-Laird random-effects model. To explore any cause of study heterogeneity, we conducted a meta-regression analysis. Results: Six studies with data from 1419 patients were included. The meta-analytic pooled inter-reader reliability of FLIS was 0.93 (95% confidence interval [CI], 0.88–0.98). That of the three FLIS subcategories were 0.93 (95% CI, 0.85–1.00), 0.95 (95% CI, 0.91–1.00), and 0.90 (95% CI, 0.81–0.99) for EnQS, ExQS, and PVsQS, respectively. The pooled FLIS data was moderately heterogenous, but heterogeneity was not associated with the study methodology, MRI-related factors, and reader experience. Conclusion: The FLIS and its three subcategories showed almost perfect inter-reader reliability. Therefore, FLIS may be a reliable imaging parameter that reflects liver function and outcomes in patients with chronic liver disease. Further studies should be conducted to confirm any factors affecting the inter-reader reliability of FLIS. Graphical abstract: [Figure not available: see fulltext.]