Using GPT-4 for LI-RADS feature extraction and categorization with multilingual free-text reports.

BACKGROUND AND AIMS: The Liver Imaging Reporting and Data System (LI-RADS) offers a standardized approach for imaging hepatocellular carcinoma. However, the diverse styles and structures of radiology reports complicate automatic data extraction. Large language models hold the potential for structured data extraction from free-text reports. Our objective was to evaluate the performance of Generative Pre-trained Transformer (GPT)-4 in extracting LI-RADS features and categories from free-text liver magnetic resonance imaging (MRI) reports. METHODS: Three radiologists generated 160 fictitious free-text liver MRI reports written in Korean and English, simulating real-world practice. Of these, 20 were used for prompt engineering, and 140 formed the internal test cohort. Seventy-two genuine reports, authored by 17 radiologists were collected and de-identified for the external test cohort. LI-RADS features were extracted using GPT-4, with a Python script calculating categories. Accuracies in each test cohort were compared. RESULTS: On the external test, the accuracy for the extraction of major LI-RADS features, which encompass size, nonrim arterial phase hyperenhancement, nonperipheral 'washout', enhancing 'capsule' and threshold growth, ranged from .92 to .99. For the rest of the LI-RADS features, the accuracy ranged from .86 to .97. For the LI-RADS category, the model showed an accuracy of .85 (95% CI: .76, .93). CONCLUSIONS: GPT-4 shows promise in extracting LI-RADS features, yet further refinement of its prompting strategy and advancements in its neural network architecture are crucial for reliable use in processing complex real-world MRI reports.

Optimizing LI-RADS: ancillary features screened from LR-3/4 categories can improve the diagnosis of HCC on MRI.

OBJECTIVE: To determine the high-efficiency ancillary features (AFs) screened from LR-3/4 lesions and the HCC/non-HCC group and the diagnostic performance of LR3/4 observations. MATERIALS AND METHODS: We retrospectively analyzed a total of 460 patients (with 473 nodules) classified into LR-3-LR-5 categories, including 311 cases of hepatocellular carcinoma (HCC), 6 cases of non-HCC malignant tumors, and 156 cases of benign lesions. Two faculty abdominal radiologists with experience in hepatic imaging reviewed and recorded the major features (MFs) and AFs of the Liver Imaging Reporting and Data System (LI-RADS). The frequency of the features and diagnostic performance were calculated with a logistic regression model. After applying the above AFs to LR-3/LR-4 observations, the sensitivity and specificity for HCC were compared. RESULTS: The average age of all patients was 54.24 ± 11.32 years, and the biochemical indicators ALT (P = 0.044), TBIL (P = 0.000), PLT (P = 0.004), AFP (P = 0.000) and Child‒Pugh class were significantly higher in the HCC group. MFs, mild-moderate T2 hyperintensity, restricted diffusion and AFs favoring HCC in addition to nodule-in-nodule appearance were common in the HCC group and LR-5 category. AFs screened from the HCC/non-HCC group (AF-HCC) were mild-moderate T2 hyperintensity, restricted diffusion, TP hypointensity, marked T2 hyperintensity and HBP isointensity (P = 0.005, < 0.001, = 0. 032, p < 0.001, = 0.013), and the AFs screened from LR-3/4 lesions (AF-LR) were restricted diffusion, mosaic architecture, fat in mass, marked T2 hyperintensity and HBP isointensity (P < 0.001, = 0.020, = 0.036, < 0.001, = 0.016), which were not exactly the same. After applying AF-HCC and AF-LR to LR-3 and LR-4 observations in HCC group and Non-HCC group, After the above grades changed, the diagnostic sensitivity for HCC were 84.96% using AF-HCC and 85.71% using AF-LR, the specificity were 89.26% using AF-HCC and 90.60% using AF-LR, which made a significant difference (P = 0.000). And the kappa value for the two methods of AF-HCC and AF-LR were 0.695, reaching a substantial agreement. CONCLUSION: When adjusting for LR-3/LR-4 lesions, the screened AFs with high diagnostic ability can be used to optimize LI-RADS v2018; among them, AF-LR is recommended for better diagnostic capabilities.

Synoptic Reporting of Focal Liver Masses in at Risk Patients: Algorithmic Diagnosis and CEUS LI-RADS.

OBJECTIVES: Reporting contrast-enhanced ultrasound (CEUS) for focal liver masses in at risk patients is a challenging task. Traditionally used prose reporting (PR) is inconsistent and lacks standardization. We propose synoptic reporting (SR), encompassing algorithmic interpretation and liver imaging and reporting data system (LI-RADS) categorization. METHODS: A software worksheet from Kailo Medical (Melbourne, AU), incorporates the CEUS algorithm for liver interpretation and CEUS LI-RADS categorization. Part 1. Feasibility of SR: twenty participants of varying experience were presented a brief lecture on SR, algorithmic approach to liver mass interpretation, and CEUS LI-RADS categorization. Ten representative liver masses were shown as unknown cases. Participants inputted data into SR worksheets. Results and LI-RADS category were generated solely by SR. Data were categorized as "correct" or "incorrect." Part 2. Prospective Analysis: Ninety-one patients for SR and 56 for PR, all were tested for completeness, efficiency, and user satisfaction. RESULTS: Part 1: Junior participants, pass rate 81.6%, and senior participants, pass rate 83.3% showed no difference in performance. Part 2: Completeness: SR 98.4% and PR 87.0%. Efficiency: Average total time to completion: SR 11 minutes and PR 20 minutes. User satisfaction: Ultrasound technologists, all referring physicians, and six out of seven radiologists preferred SR over PR. Major benefits cited were total time saved, consistency and accuracy in documentation, and report completeness. CONCLUSIONS: SR is a reliable and useful tool in clinical practice to report liver masses on ultrasound and assign an appropriate LI-RADS categorization and management pathway. This ultimately improves communication with referring clinicians and leads to better patient outcomes.

Comparing Survival Outcomes of Patients With LI-RADS-M Hepatocellular Carcinomas and Intrahepatic Cholangiocarcinomas.

BACKGROUND: There is a sparsity of data evaluating outcomes of patients with Liver Imaging Reporting and Data System (LI-RADS) (LR)-M lesions. PURPOSE: To compare overall survival (OS) and progression free survival (PFS) between hepatocellular carcinoma (HCC) and intrahepatic cholangiocarcinoma (iCCA) meeting LR-M criteria and to evaluate factors associated with prognosis. STUDY TYPE: Retrospective. SUBJECTS: Patients at risk for HCC with at least one LR-M lesion with histologic diagnosis, from 8 academic centers, yielding 120 patients with 120 LR-M lesions (84 men [mean age 62 years] and 36 women [mean age 66 years]). FIELD STRENGTH/SEQUENCE: A 1.5 and 3.0 T/3D T1 -weighted gradient echo, T2 -weighted fast spin-echo. ASSESSMENT: The imaging categorization of each lesion as LR-M was made clinically by a single radiologist at each site and patient outcome measures were collected. STATISTICAL TESTS: OS, PFS, and potential independent predictors were evaluated by Kaplan-Meier method, log-rank test, and Cox proportional hazard model. A P value of <0.05 was considered significant. RESULTS: A total of 120 patients with 120 LR-M lesions were included; on histology 65 were HCC and 55 were iCCA. There was similar median OS for patients with LR-M HCC compared to patients with iCCA (738 days vs. 769 days, P = 0.576). There were no significant differences between patients with HCC and iCCA in terms of sex (47:18 vs. 37:18, P = 0.549), age (63.0 ± 8.4 vs. 63.4 ± 7.8, P = 0.847), etiology of liver disease (P = 0.202), presence of cirrhosis (100% vs. 100%, P = 1.000), tumor size (4.73 ± 3.28 vs. 4.75 ± 2.58, P = 0.980), method of lesion histologic diagnosis (P = 0.646), and proportion of patients who underwent locoregional therapy (60.0% vs. 38.2%, P = 0.100) or surgery (134.8 ± 165.5 vs. 142.5 ± 205.6, P = 0.913). Using multivariable analysis, nonsurgical compared to surgical management (HR, 4.58), larger tumor size (HR, 1.19), and higher MELD score (HR, 1.12) were independently associated with worse OS. DATA CONCLUSION: There was similar OS in patients with LR-M HCC and LR-M iCCA, suggesting that LR-M imaging features may more closely reflect patient outcomes than histology. EVIDENCE LEVEL: 3 TECHNICAL EFFICACY: Stage 5.

Comparative Performance of 2018 LI-RADS versus Modified LIRADS (mLI-RADS): An Individual Participant Data Meta-Analysis.

BACKGROUND: LI-RADS version 2018 (v2018) is used for non-invasive diagnosis of hepatocellular carcinoma (HCC). A recently proposed modification (known as mLI-RADS) demonstrated improved sensitivity while maintaining specificity and positive predictive value (PPV) of LI-RADS category 5 (definite HCC) for HCC. However, mLI-RADS requires multicenter validation. PURPOSE: To evaluate the performance of v2018 and mLI-RADS for liver lesions in a large, heterogeneous, multi-national cohort of patients at risk for HCC. STUDY TYPE: Systematic review and meta-analysis using individual participant data (IPD) [Study Protocol: https://osf.io/duys4]. POPULATION: 2223 observations from 1817 patients (includes all LI-RADS categories; females = 448, males = 1361, not reported = 8) at elevated risk for developing HCC (based on LI-RADS population criteria) from 12 retrospective studies. FIELD STRENGTH/SEQUENCE: 1.5T and 3T; complete liver MRI with gadoxetate disodium, including axial T2w images and dynamic axial fat-suppressed T1w images precontrast and in the arterial, portal venous, transitional, and hepatobiliary phases. Diffusion-weighted imaging was used when available. ASSESSMENT: Liver observations were categorized using v2018 and mLI-RADS. The diagnostic performance of each system's category 5 (LR-5 and mLR-5) for HCC were compared. STATISTICAL TESTS: The Quality Assessment of Diagnostic Accuracy Studies version 2 (QUADAS-2 was applied to determine risk of bias and applicability. Diagnostic performances were assessed using the likelihood ratio test for sensitivity and specificity and the Wald test for PPV. The significance level was P < 0.05. RESULTS: 17% (2/12) of the studies were considered low risk of bias (244 liver observations; 164 patients). When compared to v2018, mLR-5 demonstrated higher sensitivity (61.3% vs. 46.5%, P < 0.001), similar PPV (85.3% vs. 86.3%, P = 0.89), and similar specificity (85.8% vs. 90.8%, P = 0.16) for HCC. DATA CONCLUSION: This study confirms mLR-5 has higher sensitivity than LR-5 for HCC identification, while maintaining similar PPV and specificity, validating the mLI-RADS proposal in a heterogeneous, international cohort. LEVEL OF EVIDENCE: 3 TECHNICAL EFFICACY: Stage 2.

Optimal imaging criteria and modality to determine Milan criteria for the prediction of post-transplant HCC recurrence after locoregional treatment.

OBJECTIVES: We aimed to investigate the optimal radiologic method to determine Milan criteria (MC) for the prediction of recurrence in patients who underwent locoregional treatment (LRT) for hepatocellular carcinoma (HCC) and subsequent liver transplantation (LT). METHODS: This retrospective study included 121 HCC patients who underwent LRT and had both liver dynamic CT and MRI. They were classified with MC using four cross combinations of two imaging modalities (CT and MRI) and two diagnostic criteria (modified Response Evaluation Criteria in Solid Tumors [mRECIST] and Liver Imaging Reporting and Data System treatment response algorithm [LI-RADS TRA]). Competing risk regression was performed to analyze the time to recurrence after LT. The predictive abilities of the four methods for recurrence were evaluated using the time-dependent area under the curve (AUC). RESULTS: Competing risk regression analyses found that beyond MC determined by MRI with mRECIST was independently associated with recurrence (hazard ratio, 6.926; p = 0.001). With mRECIST, MRI showed significantly higher AUCs than CT at 3 years and 5 years after LT (0.597 vs. 0.756, p = 0.012 at 3 years; and 0.588 vs. 0.733, p = 0.024 at 5 years). Using the pathologic reference standard, MRI with LI-RADS TRA showed higher sensitivity (61.5%) than CT with LI-RADS TRA (30.8%, p < 0.001) or MRI with mRECIST (38.5%, p < 0.001). CONCLUSIONS: MRI with mRECIST was the optimal radiologic method to determine MC for the prediction of post-LT recurrence in HCC patients with prior LRT. KEY POINTS: • MRI with modified RECIST (mRECIST) is the optimal preoperative method to determine Milan criteria for the prediction of post-transplant HCC recurrence in patients with prior locoregional treatment. • With mRECIST, MRI was better than CT for the prediction of post-transplant recurrence.

Deep learning-assisted LI-RADS grading and distinguishing hepatocellular carcinoma (HCC) from non-HCC based on multiphase CT: a two-center study.

OBJECTIVES: To develop a deep learning (DL) method that can determine the Liver Imaging Reporting and Data System (LI-RADS) grading of high-risk liver lesions and distinguish hepatocellular carcinoma (HCC) from non-HCC based on multiphase CT. METHODS: This retrospective study included 1049 patients with 1082 lesions from two independent hospitals that were pathologically confirmed as HCC or non-HCC. All patients underwent a four-phase CT imaging protocol. All lesions were graded (LR 4/5/M) by radiologists and divided into an internal (n = 886) and external cohort (n = 196) based on the examination date. In the internal cohort, Swin-Transformer based on different CT protocols were trained and tested for their ability to LI-RADS grading and distinguish HCC from non-HCC, and then validated in the external cohort. We further developed a combined model with the optimal protocol and clinical information for distinguishing HCC from non-HCC. RESULTS: In the test and external validation cohorts, the three-phase protocol without pre-contrast showed κ values of 0.6094 and 0.4845 for LI-RADS grading, and its accuracy was 0.8371 and 0.8061, while the accuracy of the radiologist was 0.8596 and 0.8622, respectively. The AUCs in distinguishing HCC from non-HCC were 0.865 and 0.715 in the test and external validation cohorts, while those of the combined model were 0.887 and 0.808. CONCLUSION: The Swin-Transformer based on three-phase CT protocol without pre-contrast could feasibly simplify LI-RADS grading and distinguish HCC from non-HCC. Furthermore, the DL model have the potential in accurately distinguishing HCC from non-HCC using imaging and highly characteristic clinical data as inputs. CLINICAL RELEVANCE STATEMENT: The application of deep learning model for multiphase CT has proven to improve the clinical applicability of the Liver Imaging Reporting and Data System and provide support to optimize the management of patients with liver diseases. KEY POINTS: • Deep learning (DL) simplifies LI-RADS grading and helps distinguish hepatocellular carcinoma (HCC) from non-HCC. • The Swin-Transformer based on the three-phase CT protocol without pre-contrast outperformed other CT protocols. • The Swin-Transformer provide help in distinguishing HCC from non-HCC by using CT and characteristic clinical information as inputs.

Comparison of the Feasibility and Diagnostic Performance of ACR CEUS LI-RADS and a Modified CEUS LI-RADS for HCC in Examinations Using Sonazoid.

OBJECTIVES: The present study aimed to determine the feasibility of the American College of Radiology's (ACR) contrast-enhanced ultrasound (CEUS) Liver Imaging Reporting and Data System (LI-RADS) (version 2017) in examinations using Sonazoid and compare its diagnostic performance with that of modified LI-RADS in patients at high risk of hepatocellular carcinoma (HCC). METHODS: This retrospective study's sample population consisted of 137 participants with a total of 140 nodules who underwent CEUS with Sonazoid and pathological confirmation via surgery or biopsy from January 2020 to February 2022. The lesions were evaluated and classified based on the reference standards (ie, ACR CEUS LI-RADS and modified LI-RADS). The overall diagnostic capabilities of the two systems were evaluated in terms of accuracy, sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) with 95% confidence intervals (CIs). RESULTS: The participants had a median age of 51 years and an interquartile range of 43-58 years. Regarding LR-5 as a predictor of HCC, the accuracy results of the ACR LI-RADS and modified LI-RADS algorithms were 72.9 and 71.4%, respectively (P = .50). The sensitivity of both systems was the same (69.7%; 95% CI: 60.7-77.8%). Regarding LR-M as a predictor of non-HCC malignancy, the diagnostic performance of the algorithms was the same, with accuracy and sensitivity results of 76.4 and 73.3%, respectively (95% CI: 44.9-92.2%). CONCLUSION: The findings indicate that modified LI-RADS had a moderate level of diagnostic performance for HCC in examinations using Sonazoid, which was comparable to ACR LI-RADS.

Non-invasive diagnosis strategy of hepatocellular carcinoma in low-risk population.

AIMS: With prevalence of hepatocellular carcinoma (HCC) in low-risk population (LRP), establishing a non-invasive diagnostic strategy becomes increasingly urgent to spare unnecessary biopsies in this population. The purposes of this study were to find characterisics of HCC and to establish a proper non-invasive method to diagnose HCC in LRP. METHODS: A total of 681 patients in LRP (defined as the population without cirrhosis, chronic HBV infection or HCC history) were collected from 2 institutions. The images of computed tomography (CT) and magnetic resonance imaging (MRI) were manually analysed. We divided the patients into the training cohort (n = 324) and the internal validating cohort (n = 139) by admission time in the first institution. The cohort in the second institution was viewed as the external validation (n = 218). A multivariate logistic regression model incorporating both imaging and clinical independent risk predictors was developed. C-statistics was used to evaluate the diagnostic performance. RESULTS: Besides the major imaging features of HCC (non-rim enhancement, washout and enhancing capsule), tumor necrosis or severe ischemia (TNSI) on imaging and two clinical characteristics (gender and alpha fetoprotein) were also independently associated with HCC diagnosis (all P < 0.01). A clinical model (including 3 major features, TNSI, gender and AFP) was built to diagnose HCC and achieved good diagnostic performance (area under curve values were 0.954 in the training cohort, 0.931 in the internal validation cohort and 0.902 in the external cohort). CONCLUSIONS: The clinical model in this study developed a satisfied non-invasive diagnostic performance for HCC in LRP.

Evaluation of contrast-enhanced ultrasound LI-RADS version 2017: application on 271 liver nodules in individuals with non-alcoholic steatohepatitis.

OBJECTIVES: To evaluate the diagnostic performance of contrast-enhanced ultrasound (CEUS) Liver Imaging Reporting and Data System (LI-RADS) version 2017 in a population with non-alcoholic steatohepatitis (NASH). METHODS: Between January 2013 and December 2020, consecutive patients diagnosed with NASH with untreated liver nodules were enrolled in this retrospective study. A prospective evaluation was performed between January 2021 and August 2021 as a validation set. Diagnostic performance was assessed. RESULTS: We included 217 nodules in 211 patients (mean age, 49.7 ± 21.7 years; male, 156) in the retrospective study. The positive predictive value (PPV) of CEUS LR-5 in the diagnosis of hepatocellular carcinoma (HCC) was 70.8% (56/79). In total, 28 of 217 (12.9%) nodules were classified as LR-M, of which 12 showed arterial phase hyper-enhancement, early washout, and absence of a punched-out appearance within 5 min; 10 of the 12 (83%) were HCC. When these nodules were reclassified as LR-5, the specificity of LR-M as a predictor of non-HCC malignancy increased from 91.0 (181/199) to 96.5% (192/199) (p = .023). Despite the reclassification, LR-5 specificity and PPV remained high (80.6% and 72.5%, respectively). Following reclassification, LR-M specificity increased from 90.0 (45/50) to 100% (50/50) (p = .022) in the validation set. CONCLUSION: CEUS LI-RADS category LR-5 is effective in predicting the presence of HCC. In NASH patients, diagnostic performance can be further improved by reclassifying LR-M nodules with arterial phase hyper-enhancement, early washout, and punched-out appearance as LR-5. KEY POINTS: • The LI-RADS classification of CEUS has a high application value for differentiation of HCC in NASH patients. • When LR-M nodules with arterial phase hyperenhancement and early washout but not punched-out appearance at < 5 min are reclassified as LR-5; the modification of LI-RADS has a better performance. • The PPV of modified LR-5 in the non-cirrhotic group was better than that of LR-5. The PPV of modified CEUS LR-5 in the non-cirrhotic group was comparable to that in the cirrhotic group (p both = 0.065).

What proportion of LI-RADS 5 observations reported in clinical practice do not meet LI-RADS 5 criteria?

OBJECTIVES: Liver Imaging Reporting and Data System (LI-RADS, LR) category 5 (definite hepatocellular carcinoma [HCC]) is assigned based on combinations of major imaging features (MFs): size, arterial-phase hyperenhancement (APHE), washout (WO), enhancing capsule, and threshold growth. The criteria were simplified in v2018 compared to v2017. The goal of this study is to assess the proportion of LR-5 observations reported in clinical practice with LI-RADS v2017 or v2018 that did not meet LR-5 criteria based on reported MFs. METHODS: All MR and CT reports using a standardized LI-RADS template between April 2017 and September 2020 were identified retrospectively. For each reported LR-5 observation, size, MFs, and LI-RADS version (v2017 or v2018) were extracted. Reported MFs were used to determine whether LR-5 criteria were met using the applied version of LI-RADS. The data was summarized descriptively. RESULTS: Three hundred eight observations in 234 patients (67.6% male, mean age 66.2 years) were reported as LR-5, including 136 (44.2%) with v2017 and 172 (55.8%) with v2018. 8/136 (6%) v2017 LR-5 observations and 6/172 (3%) v2018 LR-5 observations did not meet LR-5 criteria. Of 8 incorrectly categorized v2017 observations, 3 (43%) lacked APHE, 1 (14%) was a 16-mm new observation with APHE only, and 4 (43%) were 10-19 mm with APHE and WO. Of the 6 incorrectly categorized v2018 observations, 5 (83%) lacked APHE and 1 (17%) was < 10 mm. CONCLUSIONS: Depending on the LI-RADS version, 3-6% of LR-5 observations reported in clinical practice do not meet LR-5 criteria based on reported MFs. Key Points • Depending on the LI-RADS version, 3-6% of LR-5 observations in clinical practice do not meet LR-5 criteria based on reported major imaging features. • Assigning LR-5 category to observations without nonrim arterial-phase hyperenhancement was the most common error.

Imaging Features of Hepatocellular Carcinoma in Children With and Without Underlying Risk Factors.

BACKGROUND. Pediatric hepatocellular carcinoma (HCC) is an aggressive malignancy for which imaging findings remain poorly described. In comparison with adult HCC, pediatric HCC more commonly occurs without underlying risk factors, and standardized surveillance guidelines for those with predispositions are lacking. OBJECTIVE. The purpose of this article was to evaluate imaging findings of nonfibrolamellar pediatric HCC and to identify associations between these imaging findings and the presence of predisposing factors. METHODS. This retrospective study included children (≤ 18 years) with histologically confirmed nonfibrolamellar HCC who underwent multiphase CT or MRI at one of four academic children's hospitals between July 2009 and April 2019. Surveillance regimens in children with predispositions were at the discretion of treating physicians. Clinical characteristics were recorded. Scan indications were classified as surveillance versus clinical signs and symptoms. Images from all sites were submitted to a cloud-based server. Two radiologists independently assessed imaging features of HCC, including tumor size, tumor in vein, Pre-Treatment Extent of Tumor (PRETEXT) stage, and LI-RADS major features of adult HCC. Imaging findings were compared between patients with and without predispositions. RESULTS. The study included 39 patients: 17 with predispositions (mean age, 10.5 ± 4.5 years; nine boys, eight girls) and 22 without predispositions (mean age, 11.3 ± 5.1 years; 12 boys, 10 girls). Scan indication was surveillance in 14/17 patients with predispositions versus 0/22 patients without predispositions (p < .001). Patients with versus those without predispositions had smaller tumor size (reader 1: 6.0 vs 11.9 cm [p = .003]; reader 2: 6.0 vs 12.9 cm [p < .001]) and less frequent tumor in vein (reader 1: 0% vs 41% [p = .002]; reader 2: 0% vs 36% [p = .006]). PRETEXT stage IV disease was observed in 18% (both readers) of patients with predispositions versus 50-55% of patients without predispositions. No LI-RADS major feature of adult HCC showed a significant difference in frequency between patients with and without predispositions for either reader (all p > .05). CONCLUSION. Among children with HCC, those with predispositions exhibited smaller and lower-stage tumors and less frequent tumor in vein, likely because of surveillance imaging. CLINICAL IMPACT. The study supports the role of routine surveillance imaging in children with HCC predispositions to facilitate earlier detection. Standardization of surveillance guidelines remains needed.

Contrast-Enhanced Ultrasound Using Perfluorobutane: Impact of Proposed Modified LI-RADS Criteria on Hepatocellular Carcinoma Detection.

BACKGROUND. Contrast-enhanced ultrasound (CEUS) LI-RADS version 2017 (v2017) applies only to CEUS examinations performed using pure blood pool agents, noting that future versions will address combined blood pool and Kupffer cell agents such as perfluorobutane. Such agents may improve hepatocellular carcinoma (HCC) detection by visualization of a defect in the Kupffer phase (obtained ≥ 10 minutes after injection). OBJECTIVE. The purpose of our study was to compare the diagnostic performance of the LR-5 category for HCC detection in at-risk patients between CEUS LI-RADS v2017 and proposed modified criteria for CEUS examinations performed using perfluorobutane. METHODS. This retrospective study included 293 patients at risk for HCC (259 men, 34 women; mean age, 55 ± 12 [SD] years) who underwent CEUS using perfluorobutane from March 1, 2020, to October 30, 2020, showing a total of 304 observations (274 HCC, 14 non-HCC malignancy, and 16 benign lesions). Two readers independently assessed examinations and assigned categories using both CEUS LI-RADS v2017 and the proposed modified criteria. In the modified criteria, observations 10 mm or greater with not rim arterial phase hyperenhancement (APHE), no washout, and a Kupffer defect were upgraded from LR-4 to LR-5, and observations 10 mm or greater with not rim APHE, early washout, and a mild Kupffer defect were reassigned from LR-M to LR-5. Interreader agreement was assessed, and consensus interpretations were reached. Diagnostic performance was evaluated. RESULTS. Interreader agreement for LI-RADS category assignments, expressed using kappa coefficients, was 0.839 for CEUS LI-RADS v2017 and 0.854 for the modified criteria. Modified criteria upgraded 35 observations from LR-4 to LR-5 on the basis of a Kupffer defect, of which 34 were HCC and one was benign. Modified criteria reassigned 22 observations from LR-M to LR-5 on the basis of a mild Kupffer defect, of which all were HCC. LR-5 using modified criteria, compared with CEUS LI-RADS v2017, had significantly increased sensitivity (89% vs 69%, p < .001), a nonsignificant decrease in specificity (83% vs 87%, p > .99), and significantly increased accuracy (89% vs 71%, p < .001) for HCC. CONCLUSION. When using perfluorobutane for CEUS in at-risk patients, modified criteria incorporating Kupffer defects significantly improve sensitivity without significant loss of specificity in HCC detection. CLINICAL IMPACT. Future CEUS LI-RADS updates seeking to address the use of combined blood pool and Kupffer cell agents should consider adoption of the explored criteria.

Outcomes of LI-RADS US-2 Subthreshold Observations Detected on Surveillance Ultrasound.

BACKGROUND. Ultrasound LI-RADS version 2017 recommends that patients with US-2 subthreshold observations undergo repeat surveillance ultrasound in 3-6 months and return to routine surveillance if the observation shows no growth for 2 years. However, outcomes of US-2 observations are unknown. OBJECTIVE. The purpose of this article was to determine imaging outcomes of US-2 observations detected on surveillance ultrasound examinations. METHODS. This retrospective study included 175 patients (median age, 59 years; 70 women, 105 men) at high risk for hepatocellular carcinoma (HCC) with US-2 observations (i.e., subcentimeter observations) on surveillance ultrasound. Observations were classified on follow-up ultrasound performed 2 or more years later as showing no correlate, stable (if remaining subcentimeter), or progressed (if measuring ≥ 10 mm, meeting US-3 criteria). Observations were classified on follow-up multiphasic CT or MRI (stratified as < 2-year vs ≥ 2-year follow-up) as showing no correlate or, if showing a correlate, using CT/MRI LI-RADS version 2018. RESULTS. A total of 111 patients had follow-up ultrasound after 2 or more years and 106 had follow-up CT or MRI (79 before 2 years, 27 after 2 years). On the basis of final follow-up examinations, 173 of 175 observations were stable on follow-up ultrasound 2 or more years later (n = 68); showed no correlate on follow-up ultrasound, CT, or MRI (n = 88); or were classified as LR-1 or LR-2 on CT or MRI (n = 17). The remaining 2 of 175 observations were LR-3 on CT or MRI. No observations progressed to US-3 on follow-up ultrasound or were classified as LR-4 or greater on CT or MRI. A correlate was observed in 25 of the 106 follow-up CT or MRI examinations (LR-1 or LR-2 in 23; LR-3 in two). Eight patients developed HCC at a median of 2.0 years after initial US-2 observation detection; all HCCs were in separate locations from the baseline observations and were preceded by a surveillance ultrasound that could not reidentify the baseline observation. In three patients who underwent liver transplant, the explant showed no dysplastic nodule or HCC. CONCLUSION. US-2 subthreshold observations are unlikely to progress or become HCC and commonly have no correlate on follow-up imaging. CLINICAL IMPACT. Because of the low progression rate of US-2 subthreshold observations, it is unclear if an extended period of intensive surveillance, as recommended by multiple professional societies, is warranted.

Associations of Ultrasound LI-RADS Visualization Score With Examination, Sonographer, and Radiologist Factors: Retrospective Assessment in Over 10,000 Examinations.

BACKGROUND. When performing ultrasound (US) for hepatocellular carcinoma (HCC) screening, numerous factors may impair hepatic visualization, potentially lowering sensitivity. US LI-RADS includes a visualization score as a technical adequacy measure. OBJECTIVE. The purpose of this article is to identify associations between examination, sonographer, and radiologist factors and the visualization score in liver US HCC screening. METHODS. This retrospective study included 6598 patients (3979 men, 2619 women; mean age, 58 years) at risk for HCC who underwent a total of 10,589 liver US examinations performed by 91 sonographers and interpreted by 50 radiologists. Visualization scores (A, no or minimal limitations; B, moderate limitations; C, severe limitations) were extracted from clinical reports. Patient location (emergency department [ED], in-patient, outpatient), sonographer and radiologist liver US volumes during the study period (< 50, 50-500, > 500 examinations), and radiologist practice pattern (US, abdominal, community, interventional) were recorded. Associations with visualization scores were explored. RESULTS. Frequencies of visualization scores were 71.5%, 24.2%, and 4.2% for A, B, and C, respectively. Scores varied significantly (p < .001) between examinations performed in ED patients (49.8%, 40.1%, and 10.2%), inpatients (58.8%, 33.9%, and 7.3%), and outpatients (76.7%, 20.3%, and 2.9%). Scores also varied significantly (p < .001) by sonographer volume (< 50 examinations: 58.4%, 33.7%, and 7.9%; > 500 examinations: 72.9%, 22.5%, and 4.6%); reader volume (< 50 examinations: 62.9%, 29.9%, and 7.1%; > 500 examinations: 67.3%, 28.0%, and 4.7%); and reader practice pattern (US: 74.5%, 21.3%, and 4.3%; abdominal: 67.0%, 28.1%, and 4.8%; community: 75.2%, 21.9%, and 2.9%; interventional: 68.5%, 24.1%, and 7.4%). In multivariable analysis, independent predictors of score C were patient location (ED/inpatient: odds ratio [OR], 2.62; p < .001) and sonographer volume (< 50: OR, 1.55; p = .01). Among sonographers performing 50 or more examinations, the percentage of outpatient examinations with score C ranged from 0.8% to 5.4%; 9/33 were above the upper 95% CI of 3.2%. CONCLUSION. The US LI-RADS visualization score may identify factors affecting quality of HCC screening examinations and identify outlier sonographers in terms of poor examination quality. The approach also highlights potential systematic biases among radiologists in their quality assessment process. CLINICAL IMPACT. These findings may be applied to guide targeted quality improvement efforts and establish best practices and performance standards for screening programs.

Multicenter Study of ACR Ultrasound LI-RADS Visualization Scores on Serial Examinations: Implications for Surveillance Strategies.

BACKGROUND. The American College of Radiology Ultrasound LI-RADS includes the visualization score as a subjective measure of examination quality and the expected level of sensitivity. Whether a single suboptimal visualization score warrants a change in surveillance strategy is unknown. OBJECTIVE. The purpose of this study is to determine the relative stability of visualization scores on serial surveillance ultrasound examinations in patients at risk for HCC. METHODS. This retrospective study included patients at risk for HCC who underwent at least two HCC surveillance ultrasound examinations at one of three institutions between January 2017 and November 2020. The frequencies of the score remaining unchanged after variable numbers of preceding examinations with the given score were determined. A mixed-effects logistic model was fitted to identify factors associated with a repeat score of C (denoting severe limitations) versus a change to score A (indicating no or minimal limitations) or score B (denoting moderate limitations). RESULTS. A total of 3169 patients underwent at least two ultrasound examinations, yielding a total of 9602 examinations. A total of 8030 examinations (83.6%) were assigned score A; 1378 (14.4%), score B; and 194 (2.0%), score C. The frequency of score A was 88%, 91%, and 93% after one, two, and three consecutive prior examinations with score A, respectively. The frequency of score B was 45%, 48%, and 55% after one, two, and three consecutive prior examinations with score B. The frequency of score C was 42%, 67%, and 80% after one, two, and three consecutive prior examinations with score C. Among 109 examinations with score C in 91 patients with an available follow-up examination, no factor (including age, sex, severe steatosis, advanced cirrhosis, ascites, body mass index, and a change in the ultrasound machine, sonographer, or radiologist) was significantly associated with repeat score C (all p > .05). Although not statistically significant, presence of severe steatosis and presence of advanced cirrhosis had the highest odds ratios (2.88 and 2.38, respectively) for repeat score C in multivariable analysis. CONCLUSION. Only 42% of patients with visualization score C on a single surveillance examinations have score C on follow-up examinations; however, after two or more score C examinations, the chance of future score C substantially increases. CLINICAL IMPACT. The findings may inform decisions regarding alternative surveillance strategies in patients with visualization score C on ultrasound. This decision should consider the number of previous examinations with score C.

LI-RADS Version 2018 Targetoid Appearances on Gadoxetic Acid-Enhanced MRI: Interobserver Agreement and Diagnostic Performance for the Differentiation of HCC and Non-HCC Malignancy.

BACKGROUND. In LI-RADS version 2018, observations showing at least one of five targetoid appearances in different sequences or postcontrast phases are categorized LR-M, indicating likely non-hepatocellular carcinoma (HCC) malignancy. OBJECTIVE. The purpose of this study was to evaluate interobserver agreement for LI-RADS targetoid appearances among a large number of radiologists of varying experience and the diagnostic performance of targetoid appearances for differentiating HCC from non-HCC malignancy. METHODS. This retrospective study included 100 patients (76 men, 24 women; mean age, 58 ± 9 [SD] years) at high risk of HCC who underwent gadoxetic acid-enhanced MRI within 30 days before hepatic tumor resection (25 randomly included patients with non-HCC malignancy [13, intrahepatic cholangiocarcinoma; 12, combined HCC-cholangiocarcinoma]; 75 matched patients with HCC). Eight radiologists (four more experienced [8-15 years]; four less experienced [1-5 years]) from seven institutions independently assessed observations for the five targetoid appearances and LI-RADS categorization. Interobserver agreement and diagnostic performance for non-HCC malignancy were evaluated. RESULTS. Interobserver agreement was poor for peripheral washout (κ = 0.20); moderate for targetoid transitional phase or hepatobiliary phase appearance (κ = 0.33), delayed central enhancement (κ = 0.37), and targetoid restriction (κ = 0.43); and substantial for rim arterial phase hyperenhancement (κ = 0.61). Agreement was fair for at least one targetoid appearance (κ = 0.36) and moderate for at least two, three, or four targetoid appearances (κ = 0.43-0.51). Agreement for individual targetoid appearances was not significantly different between more experienced and less experienced readers other than for targetoid restriction (κ = 0.63 vs 0.43; p = .001). Agreement for at least one targetoid appearance was fair among more experienced (κ = 0.29) and less experienced (κ = 0.37) reviewers. Agreement for at least two, three, or four targetoid appearances was moderate to substantial among more experienced reviewers (κ = 0.45-0.63) and moderate among less experienced reviewers (κ = 0.42-0.56). Existing LR-M criteria of at least one targetoid appearance had median accuracy for non-HCC malignancy of 62%, sensitivity of 84%, and specificity of 54%. For all reviewers, accuracy was highest when at least three (median accuracy, 79%; sensitivity, 68%; specificity, 82%) or four (median accuracy, 80%; sensitivity, 54%; specificity, 88%) targetoid appearances were required. CONCLUSION. Targetoid appearances and LR-M categorization exhibited considerable interobserver variation among both more and less experienced reviewers. CLINICAL IMPACT. Requiring multiple targetoid appearances for LR-M categorization improved interobserver agreement and diagnostic accuracy for non-HCC malignancy.

Progression Rates of LR-2 and LR-3 Observations on MRI to Higher LI-RADS Categories in Patients at High Risk of Hepatocellular Carcinoma: A Retrospective Study.

BACKGROUND. Reported rates of hepatocellular carcinoma (HCC) for LR-2 and LR-3 observations are generally greater than those expected on the basis of clinical experience, possibly reflecting some studies' requirement for pathologic reference. OBJECTIVE. The purpose of this study was to determine rates of progression to higher LI-RADS categories of LR-2 and LR-3 observations in patients at high risk of HCC. METHODS. This retrospective study included 91 patients (64 men, 27 women; mean age, 62 years) at high risk of HCC who had clinically reported LR-2 (n = 55) or LR-3 (n = 36) observations on MRI who also underwent follow-up CT or MRI at least 12 months after the observation was made. A study coordinator annotated the location of a single LR-2 or LR-3 observation per patient on the basis of the clinical reports. Using LI-RADS version 2018 criteria, two radiologists independently assigned LI-RADS categories on the follow-up examinations. Progression rates from LR-2 or LR-3 to higher categories were determined. A post hoc consensus review was performed of observations that progressed to LR-4 or LR-5. Subgroup analyses were performed with respect to presence of prior HCC (n = 34) or a separate baseline LR-5 observation (n = 12). RESULTS. For LR-2 observations, the rate of progression to LR-4 was 0.0% (95% CI, 0.0-6.7%) and to LR-5 was 3.6% (95% CI, 0.4-13.1%) for both readers. For LR-3 observations, the rate of progression to LR-4 was 22.2% (95% CI, 9.6-43.8%) and to LR-5 was 11.1% (95% CI, 3.0-28.4%) for both readers. Fourteen observations progressed to LR-4 or LR-5 for both readers. Post hoc analysis revealed no instances of progression from LR-2 to LR-4; two, from LR-2 to LR-5; eight, from LR-3 to LR-4; and four, from LR-3 to LR-5. The progression rate from LR-3 to LR-5 was higher (p < .001) among patients with (100.0%) than those without (3.0%) a separate baseline LR-5 observation for both readers. The progression rate from LR-2 to LR-5 was not associated with a separate baseline LR-5 observation for either reader (p = .30). Progression rates were not different (p > .05) between patients with versus those without prior HCC. CONCLUSION. On the basis of progression to LR-4 or LR-5, LR-2 and LR-3 observations had lower progression rates than reported in studies incorporating pathology results in determining progression. CLINICAL IMPACT. The findings refine understanding of the clinical significance of LR-2 and LR-3 observations.

Comparison of modified CEUS LI-RADS with sonazoid and CT/MRI LI-RADS for diagnosis of hepatocellular carcinoma.

AIM: To compare the diagnostic performance based on the modified CEUS Liver Imaging Reporting and Data System (LI-RADS), which includes Kupffer-phase findings as a major imaging feature, with that of CT and MRI (CT/MRI) LI-RADS for liver nodules in patients at high risk of HCC. METHODS: A total of 120 patients with 120 nodules were included in this retrospective study. The median size of the lesions was 20.0 mm (interquartile range, 14.0-30.8 mm). Of these lesions, 90.0% (108 of 120) were confirmed as HCCs, 6.7% (8 of 120) were intrahepatic cholangiocarcinomas, 1.7% (2 of 120) were metastases, and 1.7% (2 of 120) were dysplastic nodules. All nodules were diagnosed histopathologically. Each nodule was categorized according to the modified CEUS LI-RADS and CT/MRI LI-RADS version 2018. The diagnostic performance and inter-modality agreement of each criterion was compared. RESULTS: The inter-modality agreement for the modified CEUS LI-RADS and CT/MRI LI-RADS was slight agreement (kappa = 0.139, p = 0.015). The diagnostic accuracies of HCCs for the modified CEUS LR-5 and CT/MRI LR-5 were 70.0% (95% confidence interval [CI]: 61.0%, 78.0%) versus 70.8% (95% CI: 61.8%, 78.8%) (p = 0.876), respectively. The diagnostic accuracies of non-HCC malignancies for the modified CEUS LR-M and CT/MRI LR-M were 84.2% (95% CI: 76.4%, 90.2%) versus 96.7% (95% CI: 91.7%, 99.1%) (p = 0.002), respectively. CONCLUSIONS: The diagnostic performance for HCCs on the modified CEUS LR-5 and CT/MRI LR-5 are comparable. In contrast, CT/MRI LR-M has better diagnostic performance for non-HCC malignancy than that of the modified CEUS LR-M.

LR-3 and LR-4 Lesions Are More Likely to Be Hepatocellular Carcinoma in Transplant Patients with LR-5 or LR-TR Lesions.

BACKGROUND: Liver Imaging Reporting and Data System (LI-RADS) classifies liver nodules from LR-1 to LR-5 based on risk for hepatocellular carcinoma (HCC). It is challenging to know the nature of the LR-3 and LR-4 lesions. AIMS: To test our hypothesis that in patients with a definite HCC (LR-5) or treated HCC (LR-TR), a coexisting LR-3 or LR-4 lesion is more likely to represent HCC compared to patients without LR-5 or LR-TR lesions. METHODS: We conducted a retrospective study including all adult patients who received liver transplantation in our institution from 1/1/2014 to 3/3/2020 who had any LR-3 or LR-4 lesion on pre-transplant MRI. RESULTS: Seventy-eight patients were included in the final cohort (115 LR-3 and LR-4 lesions total). When accompanied by LR-5 or LR-TR lesions, 41% (28/69) of LR-3 lesions were HCC compared to 12% (3/25) when not accompanied by LR-5 LR-TR lesions. When accompanied by LR-5 or LR-TR lesions, 83% (10/12) of LR-4 lesions were HCC, versus 33% (3/9) when not accompanied by LR-5 or LR-TR lesions. In a multivariable analysis of all lesions, the presence of a LR-5 or LR-TR lesion was significantly associated with LR-3 or LR-4 lesions representing HCC (OR 6.4, p = 0.01). CONCLUSION: LR-3 and LR-4 lesions are more likely to be HCC in patients with LR-5 or LR-TR lesions. The presence of coexisting definite HCC may be a useful diagnostic feature to improve risk stratification of lesions without typical imaging features of HCC. This may also affect decision-making prior to liver transplant when HCC burden must be accurately determined.