Liver Imaging Reporting and Data System
LI-RADS is a standardized framework, maintained by the ACR, for describing, categorizing, and reporting liver findings in people at elevated risk of hepatocellular carcinoma (HCC). Rather than one rulebook, it is a family of independently versioned algorithms for different jobs and imaging methods: diagnosing untreated lesions on CT/MRI and on contrast-enhanced ultrasound, screening with grayscale ultrasound, and judging whether a treated tumor still has living tissue. Because HCC can often be diagnosed by imaging alone in at-risk patients, these algorithms feed directly into clinical and transplant decisions — and the tracks evolve on their own timelines, so different modalities sit at different version years.
The current diagnostic algorithm for untreated observations on multiphase CT and MRI; harmonized LI-RADS with AASLD guidance and refined thresholds such as the definition of threshold growth and the size/enhancement rule contributing to the definite-HCC category. Not superseded as of mid-2026.
Prior diagnostic release; brought major structural alignment and expanded the treatment-response and tumor-in-vein concepts before the 2018 refinements.
Early iteration in the original CT/MRI lineage, expanding the descriptor lexicon and category logic.
Early refinement of the original system.
The first LI-RADS release, establishing the LR-1 through LR-5 ordinal concept for liver observations.
Split the single 2017 treatment-response algorithm into two pathways because tumors look different after radiation-based therapies (radioembolization, stereotactic radiotherapy) than after non-radiation therapies (ablation, chemoembolization). The non-radiation pathway centers viability on mass-like enhancement with optional MRI ancillary features for equivocal lesions; the radiation pathway recognizes treated lesions can evolve slowly without meaning failure.
The original single treatment-response algorithm (viable / equivocal / nonviable), applied uniformly across therapy types; later judged too coarse for radiation-based therapies, prompting the 2024 split.
Refreshed the grayscale-ultrasound screening track to align with 2023 AASLD HCC guidance; folded serum AFP trend and the severe-limitation visualization score into management advice and clarified follow-up after non-positive exams. Reported to raise sensitivity at some cost to specificity.
First standardized grayscale-ultrasound surveillance track, introducing the US category and visualization-score concepts; superseded by US Surveillance v2024.
The current contrast-enhanced-ultrasound diagnostic algorithm (a revision of the 2016 release); categorizes untreated liver observations using contrast kinetics specific to microbubble agents. Notably NOT renumbered to 2024 — only the CEUS treatment-response track was added in 2024.
A new CEUS treatment-response algorithm for post-non-radiation locoregional therapy (ablation, chemo/bland embolization), assessing residual living tumor by enhancement with viable / equivocal / nonviable readouts — distinct from the older CEUS diagnostic algorithm.
The first official ACR CEUS LI-RADS; quickly revised to v2017.
The diagnostic tracks (CT/MRI and CEUS) place each liver observation on an ordinal ladder from LR-1 (almost certainly benign) up to LR-5 (almost certainly HCC), with intermediate steps, plus a separate LR-M flag for malignancy that doesn't look HCC-specific and an LR-TIV flag for tumor that has invaded a vein; LR-NC covers observations that can't be categorized. The treatment-response tracks instead report whether a treated lesion appears viable, equivocal, or nonviable. The surveillance track uses a small set of result categories paired with a separate score for how well the liver could be seen. Consult the official ACR LI-RADS documents for the exact criteria, thresholds, and decision rules.
These are our plain-language summaries. For the exact criteria, thresholds, and management rules, see the official source.
On MRI, LI-RADS LR-5 had higher PPV for HCC than KLCA-NCC definite HCC (93.9% vs 91.7%, p<0.001); on CT, PPVs were equal (92.9%). Corresponding probability categories should not be used interchangeably.
In a retrospective study of 166 hypervascular LI-RADS 3 observations <2 cm without ancillary features, only 3.0% (95% CI 0–6.1%) progressed to HCC at 1 year, far lower than the ~one-third rate often cited for the overall LR-3 category.
In retrospective study of 167 HCC lesions on MRI, moderate-to-severe hepatic iron overload independently reduced LR-5 assignment (OR 0.08, 95% CI 0.02–0.28) and altered key LI-RADS features: washout and capsule.
Background Whether the diagnostic accuracy of the Liver Imaging Reporting and Data System (LI-RADS) Treatment Response Algorithm (TRA) differs according to the type of MRI contrast agent has been unclear. Purpose To compare the diagnostic accuracy of LI-RADS Nonradiation TRA v...
LI-RADS overlaps with other HCC pathways but is more granular. The transplant-allocation system used by OPTN/UNOS defines HCC eligibility by size/enhancement classes, and LR-5 is designed to map cleanly onto that 'definite HCC' tier while adding intermediate categories the allocation system lacks. AASLD guidance incorporates LI-RADS terminology, whereas Europe's EASL and Asia's APASL/KLCA frameworks use their own diagnostic wording and sometimes accept slightly different imaging combinations as diagnostic. Compared with simply calling a lesion 'suspicious,' LI-RADS' explicit ordinal scale is what enables probability-calibrated management and research.
Across pooled analyses, the definite-HCC category (LR-5) consistently shows high specificity and strong positive predictive value, with more moderate sensitivity because many true HCCs land in lower categories — a deliberate trade-off to avoid over-calling. Diagnostic meta-analyses of CT/MRI LI-RADS report good overall performance but note inter-reader variability concentrated in the intermediate LR-3/LR-4 tiers; for CEUS, pooled LR-5 specificity is excellent with lower sensitivity. The system is widely adopted in North American practice and embedded in AASLD guidance. Common critiques are the limited standalone sensitivity of LR-5, ambiguity of the indeterminate categories, and the operational burden of multiple separately versioned algorithms.
RadPigeon is an independent radiology news digest and is not affiliated with or endorsed by American College of Radiology (ACR). “LI-RADS” is a trademark of its owner and is named here only to refer to the system. Always consult the official source for the exact, current criteria.
