Quantitative MRI biomarkers such as proton density fat fraction (PDFF) and R2* enable non-invasive assessment of liver steatosis and iron overload. Automated whole-liver segmentation can make these measurements more comprehensive and reproducible, but whether hepatic vessels should be excluded from the analysis has remained unclear.

This prospective multicenter study addressed this question in 377 patients with chronic liver disease, comparing automated whole-liver MRI analysis with and without major hepatic vessels. Histological grading from liver biopsy served as the reference standard.

The impact was remarkably small. PDFF and R2* showed very strong agreement between both segmentation approaches (ρ ≥ 0.99), with biases of only −0.06% for PDFF and −0.25 s⁻¹ for R2* when vessels were included. No patient changed PDFF-derived steatosis grade, while only 1.9% changed R2-derived iron grade*, all close to classification thresholds.

Importantly, this is the first study to specifically assess the impact of large intrahepatic vessels on PDFF and R2 precision and variability*.

The findings support simpler automated whole-liver quantification for routine clinical use, while showing that vessel exclusion can provide incremental precision for clinical trials and longitudinal studies where highly sensitive quantitative imaging biomarkers are required.