What type of study is this?

This is a Quantitative Study study (also classified as: Cohort Study).

September 17, 2025

Inter-observer Agreement for the Diagnosis of LI-RADS 3 and 4 Lesions on Gadoxetate MRI and Outcome Assessment

Key Points

Inter-observer agreement was poor for LR-3 lesions with a kappa of 0.27 and fair for LR-4 with a kappa of 0.43.
Malignancy developed in 29% of LR-3 lesions and 60% of LR-4 lesions, indicating significant diagnostic implications.
Gadoxetate MRI was used in a retrospective study of 111 cases, revealing critical outcomes linked to LR-3 and LR-4 lesions.
The findings underscore the necessity for better standardization in LI-RADS criteria to improve diagnostic accuracy.

Abstract

Motivation: Despite standardization by LI-RADS v2018, inter-observer variability persists in diagnosing liver lesions, especially for indeterminate lesions (LR-3 and 4), impacting patient outcomes. Goal(s): Assess inter-observer agreement in diagnosing LR-3 and 4 lesions on gadoxetate-enhanced MRI, describe lesion outcomes, and identify malignancy-predictive imaging features. Approach: A retrospective single-center study (n=111) analyzed inter-observer agreement, outcomes, and malignancy-predicting imaging features of LR-3 and 4 lesions. Results: Inter-observer agreement was poor for LR-3 (kappa 0.27) and fair for LR-4 (kappa 0.43). Malignancy developed in 29% of LR-3 and 60% of LR-4 lesions; LR-4 lesions progressed to HCC faster (median 18 vs. 31 months, p=0.01). Impact: The study highlights the need for improved standardization and radiologist's training in applying LI-RADS criteria to enhance diagnostic accuracy.

KI fragen

Bookmark

View Full Paper