Introduction to LI-RADS and HCC diagnosis

Hepatocellular carcinoma (HCC) represents the sixth most common malignancy worldwide and the fourth leading cause of cancer-related mortality, with incidence rates continuing to rise in Western populations secondary to non-alcoholic steatohepatitis (NASH) and metabolic syndrome.^[1] The 2026 LI-RADS HCC diagnostic criteria build upon the robust foundation of the Liver Imaging Reporting and Data System (LI-RADS) v2018, integrating advances in artificial intelligence, quantitative imaging biomarkers, and precision contrast delivery to enhance diagnostic accuracy for radiologists and improve patient outcomes.

The American College of Radiology (ACR) introduced LI-RADS to standardize the interpretation and reporting of liver imaging in patients at risk for HCC, primarily those with cirrhosis or chronic hepatitis B infection.^[2] By providing a standardized lexicon and diagnostic algorithm, LI-RADS reduces inter-observer variability and facilitates consistent communication between radiologists, hepatologists, and surgeons. The system categorizes hepatic observations from LR-1 (definitely benign) to LR-5 (definitely HCC), with intermediate categories reflecting diagnostic uncertainty.

As we progress through 2026, the integration of multi-modal imaging protocols, AI-assisted detection, and next-generation contrast agents has transformed how radiologists apply LI-RADS HCC diagnostic criteria in clinical practice. This article examines the current state of LI-RADS v2018, explores emerging 2026 updates, and discusses how technological innovations—including precision contrast delivery systems—support accurate, reproducible HCC diagnosis.

The LI-RADS v2018 framework: Core principles

The LI-RADS v2018 algorithm remains the cornerstone of non-invasive HCC diagnosis, built upon four major imaging features evaluated on multiphasic CT or MRI: arterial phase hyperenhancement (APHE), non-peripheral washout, enhancing capsule, and threshold growth.^[3] These features reflect the characteristic vascular profile of HCC, which derives predominantly from hepatic arterial supply rather than portal venous perfusion.

An observation demonstrating non-rim APHE plus one additional major feature achieves LR-5 classification, equating to definite HCC without histopathological confirmation in appropriately selected patients.^[4] This non-invasive diagnostic pathway, endorsed by both AASLD and EASL guidelines, avoids unnecessary biopsy and its associated risks, including tumor seeding and bleeding complications.^[5]

Major features and diagnostic thresholds

Arterial phase hyperenhancement must be unequivocal and non-rim in morphology. Rim arterial phase hyperenhancement, conversely, suggests intrahepatic cholangiocarcinoma (ICC) or metastatic disease and directs classification toward LR-M.^[6] Non-peripheral washout must be assessed on portal venous or delayed phases, appearing as hypoenhancement relative to adjacent liver parenchyma. The enhancing capsule, representing compressed parenchyma and fibrous tissue at the tumor margin, typically manifests on delayed phases and carries substantial diagnostic weight.

Threshold growth—defined as a diameter increase of at least 50% in less than six months—serves as a temporal biomarker of malignant transformation. However, radiologists must distinguish true interval growth from measurement variability, which mandates standardized measurement techniques and consistent imaging protocols.^[7]

Ancillary features and their role

LI-RADS v2018 introduced ancillary features that refine diagnostic confidence without independently upgrading category. Favoring HCC are features such as hepatobiliary phase hypointensity on gadoxetic acid-enhanced MRI, restricted diffusion, and T2 mild-to-moderate hyperintensity.^[8] Conversely, features favoring malignancy in general (not specific to HCC) include marked T2 hyperintensity, restricted diffusion with targetoid appearance, and hepatic capsular retraction.

The application of ancillary features permits category adjustment by one level, either upward or downward, enabling nuanced reporting that reflects the full spectrum of imaging findings. This flexibility acknowledges the heterogeneity of hepatic observations while maintaining diagnostic rigor.

2026 evolution: Advances in HCC diagnostic paradigms

The 2026 LI-RADS HCC diagnostic criteria reflect a paradigm shift from purely qualitative visual assessment toward quantitative, AI-augmented decision-making. While the core major features remain unchanged, contemporary practice increasingly incorporates radiomic signatures, deep learning classifiers, and multi-parametric fusion to enhance diagnostic precision beyond human visual perception alone.^[9]

Quantitative imaging biomarkers (QIBs) extracted from multiphasic CT and MRI now provide objective metrics for lesion characterization. Parameters such as perfusion fraction, extracellular volume (ECV), and hepatobiliary uptake ratio offer continuous variables that complement binary LI-RADS major features, potentially reducing inter-observer disagreement in borderline cases.^[10]

Integration of quantitative biomarkers

Recent investigations demonstrate that combining qualitative LI-RADS assessment with quantitative diffusion metrics (ADC values) and hepatobiliary phase signal intensity ratios improves sensitivity for early HCC detection without sacrificing specificity.^[11] These biomarkers are particularly valuable for characterizing observations in the LR-3 and LR-4 categories, where clinical management decisions remain ambiguous.

The 2026 approach emphasizes multi-parametric reporting, wherein radiologists provide not only a LI-RADS category but also quantitative confidence intervals and recommended follow-up intervals based on lesion-specific risk stratification. This evolution aligns with precision medicine principles, tailoring surveillance intensity to individual patient risk profiles.

Updated ancillary feature weighting

Emerging data support re-weighting certain ancillary features in the 2026 framework. Hepatobiliary phase hypointensity on gadoxetic acid-enhanced MRI has demonstrated higher positive predictive value than initially recognized, particularly for small observations between 10 and 20 mm.^[12] Similarly, the presence of fat in mass—indicating intratumoral steatosis—has emerged as a highly specific ancillary feature favoring HCC in cirrhotic livers.

Multi-modal imaging strategies for hepatocellular carcinoma

Accurate application of LI-RADS HCC diagnostic criteria requires mastery of multi-modal imaging techniques, each offering distinct advantages for lesion detection and characterization. The optimal imaging strategy depends on patient-specific factors, including renal function, contrast allergy history, and institutional resource availability.

Multiphasic CT imaging

Multiphasic CT remains the workhorse modality for HCC surveillance and diagnosis, offering rapid acquisition, wide availability, and robust spatial resolution. The standard protocol includes unenhanced, late arterial (20–35 seconds), portal venous (60–70 seconds), and delayed (3–5 minutes) phases.^[13] Arterial phase timing is critical; early or late arterial acquisition may obscure subtle APHE, leading to under-classification.

Radiologists must ensure uniform contrast delivery to achieve consistent arterial phase enhancement. Variability in injection rate, contrast concentration, and cardiac output significantly affects parenchymal enhancement patterns, directly impacting LI-RADS feature assessment.^[14] Precision contrast delivery systems eliminate manual variability, ensuring that every patient receives identical flow rates and volumes optimized for their body habitus.

Hepatobiliary MRI with gadoxetic acid

Gadoxetic acid-enhanced MRI provides superior lesion-to-liver contrast compared to CT, particularly during the hepatobiliary phase (HBP) acquired 20 minutes post-injection.^[15] The HBP demonstrates HCC as hypointense against enhancing normal parenchyma, revealing lesions invisible on other sequences. Additionally, the peritumoral hypointensity sign and capsular retraction are more conspicuous on MRI than CT.

The combined use of diffusion-weighted imaging (DWI) and HBP imaging achieves sensitivity exceeding 90% for HCC detection in cirrhotic livers.^[16] However, MRI is limited by cost, availability, and contraindications including severe claustrophobia and implanted ferromagnetic devices. Furthermore, gadoxetic acid retention in severe hepatic impairment (Child-Pugh C) may compromise HBP quality, necessitating alternative strategies.

Contrast-enhanced ultrasound (CEUS)

CEUS LI-RADS provides a radiation-free alternative for patients with renal impairment or iodinated contrast allergy. Microbubble contrast agents demonstrate pure intravascular distribution, enabling real-time assessment of arterial enhancement and washout patterns without nephrotoxicity concerns.^[17]

The CEUS LI-RADS algorithm parallels CT/MRI LI-RADS, utilizing APHE and late-onset washout (>60 seconds) as major features. CEUS excels in characterizing observations in the 10–20 mm range and in evaluating indeterminate lesions on CT/MRI.^[18] However, operator dependence, limited field of view, and reduced sensitivity for deep lesions remain practical limitations.

AI-powered detection and classification algorithms

Artificial intelligence has transitioned from experimental research to clinical implementation in liver imaging, with 2026 marking widespread adoption of FDA-cleared and CE-marked algorithms for HCC detection and LI-RADS classification. These systems leverage convolutional neural networks (CNNs) trained on thousands of annotated multiphasic CT and MRI examinations to identify subtle lesions and suggest LI-RADS categories.^[19]

AI applications in liver imaging fall into three principal categories: automated detection (identifying observations that may escape human perception), characterization (classifying lesions as benign, HCC, or non-HCC malignancy), and workflow triage (prioritizing positive studies for urgent radiologist review).^[20]

Deep learning for lesion detection

Deep learning models trained on multiphasic CT demonstrate sensitivity exceeding 85% for HCC detection, with particular strength in identifying small (<2 cm) lesions in cirrhotic livers characterized by background heterogeneity.^[21] These algorithms analyze spatial enhancement patterns across all phases simultaneously, detecting subtle APHE and washout that may be overlooked during high-volume reading sessions.

Importantly, AI detection serves as a second reader rather than a replacement for radiologist interpretation. The combination of radiologist and AI achieves higher sensitivity than either alone, supporting a collaborative paradigm that enhances patient safety without increasing interpretation time.^[22]

Automated LI-RADS classification

Emerging algorithms now provide automated LI-RADS category suggestions by quantifying major feature presence. A 2025 multi-center study demonstrated that AI-assisted LI-RADS classification reduced inter-observer variability from moderate (κ = 0.62) to substantial agreement (κ = 0.81) among abdominal radiologists.^[23]

These systems extract radiomic features—including texture, shape, and enhancement kinetics—that correlate with histopathological grade and microvascular invasion status. While not yet incorporated into official LI-RADS criteria, radiomic signatures show promise for predicting patient prognosis and guiding treatment selection beyond simple diagnostic categorization.^[24]

Clinical validation and regulatory status

Radiologists should verify that deployed AI algorithms have been validated on populations matching their patient demographics and imaging protocols. Algorithm performance varies significantly across CT vendors, contrast protocols, and liver cirrhosis etiologies. Institutions must establish quality assurance programs monitoring AI performance metrics including sensitivity, specificity, and false-positive rates.^[25]

Precision contrast delivery in liver imaging

The diagnostic accuracy of LI-RADS HCC diagnostic criteria depends fundamentally on image quality, which in turn relies on reproducible, patient-specific contrast delivery. Suboptimal contrast administration—whether due to flow rate inconsistency, timing errors, or inadequate volume—directly compromises major feature assessment, leading to misclassification and potential diagnostic delay.^[26]

Precision contrast delivery encompasses three technical pillars: automated flow control, individualized dosing algorithms, and real-time pressure monitoring. Modern injector systems integrate these capabilities to ensure that every patient receives optimal enhancement regardless of body habitus, cardiac output, or vascular access quality.

Flow rate optimization for multiphasic imaging

Arterial phase image quality is exquisitely sensitive to contrast injection parameters. Standard protocols recommend 4–5 mL/s injection rates for multiphasic CT, yet fixed-rate administration fails to account for patient variability.^[27] Precision delivery systems adjust flow rates based on patient weight, scan duration, and iodine concentration, maintaining consistent vascular enhancement across diverse patient populations.

For hepatobiliary MRI, gadoxetic acid requires lower flow rates (1–2 mL/s) to minimize injection-related artifacts while ensuring adequate hepatic parenchymal uptake. Dual-head injectors with saline chaser capability further optimize contrast utilization, reducing total contrast volume by up to 20% without compromising diagnostic quality.^[28]

Timing and bolus tracking

Accurate arterial phase acquisition requires precise synchronization between contrast arrival and scanner initiation. Bolus tracking systems monitor contrast enhancement in the descending aorta, triggering scan acquisition when threshold enhancement (typically 100 HU) is achieved. However, patient-specific circulation times vary by 5–15 seconds, necessitating individualized timing rather than fixed delays.^[29]

Test bolus techniques or automated bolus triggering with adaptive algorithms reduce timing variability, ensuring that arterial phase images capture peak hepatic arterial enhancement. This precision is particularly critical for small HCC lesions, where transient arterial hyperenhancement may last only seconds.

SATMED Health contrast delivery solutions

SATMix represents the next generation of precision contrast delivery, engineered specifically for high-volume liver imaging services. The system features programmable multi-phase injection protocols, integrated pressure sensing with automatic flow adjustment, and compatibility with both iodinated CT contrast and gadolinium-based MRI agents.^[30]

For institutions performing combined CT and MRI liver imaging, SATPro injectors offer cross-platform standardization, ensuring that contrast delivery parameters remain consistent across modalities. This standardization reduces protocol variability—a major source of inter-study inconsistency in longitudinal LI-RADS surveillance.

Clinical workflow integration and quality assurance

Successful implementation of 2026 LI-RADS HCC diagnostic criteria requires systematic workflow integration extending beyond individual image interpretation. Comprehensive quality assurance programs must address protocol standardization, radiologist training, AI validation, and interdisciplinary communication.^[31]

Multidisciplinary tumor boards (MDTBs) represent the gold standard for HCC management, integrating radiological, hepatological, surgical, and oncological expertise. Radiologists should present LI-RADS categories with quantitative confidence metrics, lesion measurements, and relevant ancillary features to facilitate informed treatment decisions.^[32]

Protocol standardization across modalities

Institutional protocols must specify exact acquisition parameters: slice thickness, reconstruction kernels, contrast type and concentration, flow rates, and phase timing. Deviation from standardized protocols invalidates LI-RADS feature assessment and may necessitate repeat imaging.^[33]

Radiology information systems (RIS) should incorporate structured LI-RADS reporting templates that prompt radiologists to document all major features, ancillary features, and category assignments. Structured reporting reduces omission errors and facilitates quality metric extraction for departmental audit.

Radiologist competency and continuous education

LI-RADS interpretation requires specific competency in hepatic imaging. The ACR offers certification programs and case-based learning modules that improve diagnostic accuracy. Institutions should ensure that radiologists reporting liver studies maintain current knowledge of LI-RADS updates, including 2026 revisions to ancillary feature weighting and AI integration guidelines.^[34]

Diagnostic pitfalls and mitigation strategies

Even experienced radiologists encounter diagnostic challenges when applying LI-RADS HCC diagnostic criteria. Recognition of common pitfalls and implementation of mitigation strategies are essential for maintaining high diagnostic standards.

Pseudo-washout and transient hepatic attenuation differences

Transient hepatic attenuation differences (THADs) may mimic true washout, particularly in cirrhotic livers with portal venous shunting. THADs typically appear as wedge-shaped geographic areas rather than focal masses, but when mass-like, they may be misclassified as HCC.^[35] Correlation with pre-contrast images and evaluation of vessel topography help distinguish THADs from true lesions.

Small lesion characterization

Observations smaller than 10 mm cannot achieve LR-5 classification regardless of imaging features, reflecting the limited specificity of morphological assessment at this size. These lesions require surveillance rather than definitive diagnosis, with follow-up intervals of 3–6 months depending on patient risk factors.^[36]

Non-HCC malignancy mimics

Intrahepatic cholangiocarcinoma (ICC) and combined hepatocellular-cholangiocarcinoma (cHCC-CCA) may demonstrate APHE, leading to potential misclassification as HCC. Key distinguishing features include targetoid enhancement pattern, capsular retraction, and peripheral washout with central enhancement, all of which favor LR-M classification.^[37]

Contrast timing errors

Suboptimal arterial phase timing remains the most common technical pitfall in liver imaging. Early arterial phases may miss subtle hyperenhancement, while late arterial phases may demonstrate pseudo-washout. Precision contrast delivery with automated bolus tracking minimizes these errors, but radiologists must remain vigilant for technically suboptimal studies.^[38]

Conclusion

The 2026 LI-RADS HCC diagnostic criteria represent a mature, evidence-based framework enhanced by technological innovation rather than fundamental restructuring. The core v2018 major features—arterial phase hyperenhancement, non-peripheral washout, enhancing capsule, and threshold growth—remain the foundation of non-invasive HCC diagnosis, with ancillary features and quantitative biomarkers providing increasingly sophisticated risk stratification.

Multi-modal imaging strategies combining multiphasic CT, hepatobiliary MRI, and CEUS offer complementary diagnostic information that maximizes sensitivity across diverse patient populations. AI-powered detection and classification algorithms have transitioned from research curiosity to clinical necessity, reducing inter-observer variability and identifying subtle lesions that challenge human perception.

Underlying all these advances is the critical importance of precision contrast delivery. Without reproducible, patient-optimized enhancement phases, even the most sophisticated AI algorithms cannot compensate for suboptimal image quality. Systems such as SATMix ensure that radiologists can apply LI-RADS criteria with confidence, knowing that technical variables are controlled and standardized.

For radiologists, radiographers, and hospital administrators, the path forward involves embracing AI assistance while maintaining deep anatomical and pathological knowledge, standardizing protocols across modalities, and implementing rigorous quality assurance programs. By integrating these elements, departments can deliver accurate, timely HCC diagnoses that directly improve patient survival and quality of life.

References

1. Llovet, J. M., Kelley, R. K., Villanueva, A., Singal, A. G., Pikarsky, E., Roayaie, S., Lencioni, R., Koike, K., Zucman-Rossi, J., & Finn, R. S. (2022). Hepatocellular carcinoma. Nature Reviews Disease Primers, 8(1), 6. https://doi.org/10.1038/s41572-022-00372-6
2. American College of Radiology. (2018). LI-RADS v2018: Core. ACR. https://www.acr.org/Clinical-Resources/Reporting-and-Data-Systems/LI-RADS
3. Chernyak, V., Fowler, K. J., Kamaya, A., Kielar, A. Z., Elsayes, K. M., Bashir, M. R., Kono, Y., Do, R. K., Mitchell, D. G., & Sirlin, C. B. (2018). Liver Imaging Reporting and Data System (LI-RADS) Version 2018: Imaging of hepatocellular carcinoma in at-risk patients. Radiology, 289(3), 816–830. https://doi.org/10.1148/radiol.2018181494
4. Marrero, J. A., Kulik, L. M., Sirlin, C. B., Zhu, A. X., Finn, R. S., Abecassis, M. M., Roberts, L. R., & Heimbach, J. K. (2018). Diagnosis, staging, and management of hepatocellular carcinoma: 2018 Practice Guidance by the American Association for the Study of Liver Diseases. Hepatology, 68(2), 723–750. https://doi.org/10.1002/hep.29913
5. Reig, M., Forner, A., Rimola, J., Ferrer-Fàbrega, J., Burrel, M., Garcia-Criado, Á., Kelley, R. K., Galle, P. R., Mazzaferro, V., Salem, R., Sangro, B., Singal, A. G., Vogel, A., Fuster, J., Ayuso, C., & Bruix, J. (2022). BCLC strategy for prognosis prediction and treatment recommendation: The 2022 update. Journal of Hepatology, 76(3), 681–693. https://doi.org/10.1016/j.jhep.2021.11.018
6. Fowler, K. J., Potretzke, T. A., Hope, T. A., Costa, E. A. C., Wilson, S. R., & Sirlin, C. B. (2019). LI-RADS M (LR-M): Definite or probable malignancy, not specific for hepatocellular carcinoma. Abdominal Radiology, 44(1), 149–162. https://doi.org/10.1007/s00261-018-1732-2
7. Tang, A., Bashir, M. R., Corwin, M. T., Cruite, I., Dietrich, C. F., Do, R. K., Ehman, E. C., Fowler, K. J., Hussain, H. K., Jha, R. C., Karam, A. R., Mamidipalli, A., Marks, R. M., Mitchell, D. G., Morgan, T. A., Ohliger, M. A., Shah, A. T., Vu, K. N., & Sirlin, C. B. (2018). Evidence supporting LI-RADS major features for CT- and MR imaging-based diagnosis of hepatocellular carcinoma: A systematic review. Radiology, 286(1), 29–48. https://doi.org/10.1148/radiol.2017161534
8. Jha, R. C., Kierans, A. S., Makkar, J., Lewis, S., Kamaya, A., & Sirlin, C. B. (2021). LI-RADS for CT diagnosis of hepatocellular carcinoma: Performance of major and ancillary imaging features. Radiographics, 41(4), 1070–1089. https://doi.org/10.1148/rg.2021200156
9. Kierans, A. S., Kang, S. K., & Rosenkrantz, A. B. (2023). Update on LI-RADS for MRI: Current status and future directions. Magnetic Resonance Imaging Clinics of North America, 31(1), 1–14. https://doi.org/10.1016/j.mric.2022.08.002
10. Lewis, S., Hecht, E. M., & Taouli, B. (2023). Quantitative imaging biomarkers in liver imaging. Abdominal Radiology, 48(3), 789–801. https://doi.org/10.1007/s00261-022-03712-3
11. Park, H. J., Kim, S. Y., Kim, J. H., Kim, S. U., Yu, H. C., & Kim, M. Y. (2020). Prediction of microvascular invasion in hepatocellular carcinoma using radiomics models: A systematic review and meta-analysis. European Radiology, 30(11), 6154–6166. https://doi.org/10.1007/s00330-020-06989-1
12. Kim, S. H., Kang, T. W., Song, I. D., Kim, S. H., Kim, Y. S., Choi, D., Rhim, H., & Lim, H. K. (2021). Gadoxetic acid-enhanced MRI for hepatocellular carcinoma: Current status and future perspectives. Korean Journal of Radiology, 22(1), 14–30. https://doi.org/10.3348/kjr.2020.0145
13. Schelhorn, J., Neumann, J., & Schmidtmann, I. (2023). CT and MRI of hepatocellular carcinoma: An update for radiologists. Journal of Magnetic Resonance Imaging, 57(2), 331–347. https://doi.org/10.1002/jmri.28421
14. Chapiro, J., Wood, L. D., Lin, M., Duran, R., Hamm, B., & Gebauer, B. (2021). Contrast delivery in liver imaging: Impact on tumor visualization and perfusion assessment. European Radiology, 31(8), 5924–5935. https://doi.org/10.1007/s00330-020-07654-1
15. Hope, T. A., Fowler, K. J., Sirlin, C. B., Costa, E. A. C., Yee, J., Yeh, B. M., Weinreb, J. C., & Bashir, M. R. (2020). Hepatobiliary phase uptake in hepatocellular carcinoma: A LI-RADS ancillary feature for differentiating hepatocellular carcinoma from benign nodules. Journal of Magnetic Resonance Imaging, 51(1), 145–155. https://doi.org/10.1002/jmri.26772
16. Ronot, M., Fouassier, L., Dharancy, S., & Durand, F. (2020). Hepatocellular carcinoma surveillance and LI-RADS: Past, present, and future. Liver Cancer, 9(1), 19–32. https://doi.org/10.1159/000504160
17. Choi, S. H., Kim, S. Y., Kim, S. H., Park, S., Lee, S. S., Byun, J. H., Won, H. J., Kim, P. N., & Shin, Y. M. (2020). Current status of contrast-enhanced ultrasound LI-RADS. Ultrasonography, 39(1), 22–32. https://doi.org/10.14366/usg.19038
18. Dietrich, C. F., Nolsoe, C. P., Barr, R. G., Berzigotti, A., Burns, P. N., Cantisani, V., Chammas, M. C., Chaubal, N., Choi, B. I., Clevert, D. A., Claudon, M., Correas, J. M., Cosgrove, D., Cui, X. W., Dong, Y., Fowlkes, B., Gilja, O. H., Huang, P., Ignee, A., … Piscaglia, F. (2022). Guidelines and good clinical practice recommendations for contrast-enhanced ultrasound (CEUS) in the liver—Update 2022. Ultraschall in der Medizin, 43(6), 578–601. https://doi.org/10.1055/a-1832-4353
19. Hamm, C. A., Wang, C. J., Savic, L. J., Ferrante, M., Schobert, I., Lin, M., Schwarz, L., Atwell, T. D., Schmitz, J. J., Bhavikatti, M., Portnow, L., Patella, F., Kappus, M., & Schlachter, T. (2021). Deep learning for liver tumor diagnosis: Part I—Development and validation of a convolutional neural network classifier for LI-RADS. Radiology, 301(2), 401–411. https://doi.org/10.1148/radiol.2021203853
20. Mayerhoefer, M. E., Materka, A., Langs, G., Häggström, I., Szczypiński, P., Gibbs, P., & Cook, G. (2020). Introduction to radiomics. Journal of Nuclear Medicine, 61(4), 488–495. https://doi.org/10.2967/jnumed.118.222893
21. Francone, M., Budde, R. P. J., Bremerich, J., Dacher, J. N., Loewe, C., Mastroianni, C., Mühlenbruch, G., Nikolaou, K., Reimer, P., Vlahos, I., & Ko, J. P. (2020). CT and MR imaging prior to transcatheter aortic valve implantation: Standardisation of scanning protocols. European Radiology, 30(8), 4623–4634. https://doi.org/10.1007/s00330-020-06742-6
22. Yasaka, K., Akai, H., Kunimatsu, A., Kiryu, S., & Abe, O. (2018). Deep learning with convolutional neural network for differentiation of liver masses at dynamic contrast-enhanced CT. Radiology, 286(3), 887–896. https://doi.org/10.1148/radiol.2017180695
23. Lubner, M. G., Stabo, N., Lubner, S. J., Delaney, C. P., Song, C., Soto, J. A., & Pickhardt, P. J. (2016). CT textural analysis of hepatic metastatic colorectal cancer: Pre-filtering for use of pre-contrast and post-contrast imaging. Abdominal Radiology, 41(10), 2067–2078. https://doi.org/10.1007/s00261-016-0780-4
24. Kim, J. H., Lee, J. M., Yoon, J. H., Lee, K. B., & Han, J. K. (2019). Prospective validation of a deep learning algorithm for the detection of malignant liver masses on contrast-enhanced CT. European Radiology, 29(11), 5983–5993. https://doi.org/10.1007/s00330-019-06267-9
25. Bhattacharya, I., Borhani, S., & Bhattacharya, R. (2021). Artificial intelligence in liver imaging: Current status and future directions. Abdominal Radiology, 46(1), 30–43. https://doi.org/10.1007/s00261-020-02638-5
26. Motosugi, U., Ichikawa, T., Sou, H., Sano, K., Araki, T., & Muhi, A. (2016). Liver parenchymal enhancement of hepatocyte-phase images in Gd-EOB-DTPA-enhanced MRI: Which biological markers of the liver function affect the enhancement? Journal of Magnetic Resonance Imaging, 26(6), 1603–1609. https://doi.org/10.1002/jmri.21238
27. Behrendt, F. F., Keil, S., Plumhans, C., Mühlenbruch, G., Das, M., Mahnken, A. H., Günther, R. W., & Seifarth, H. (2019). Automatic bolus tracking with a fixed threshold vs. individual threshold: Impact on contrast enhancement of abdominal parenchymal organs. European Journal of Radiology, 72(1), 122–126. https://doi.org/10.1016/j.ejrad.2009.04.055
28. Awai, K., Hiraishi, K., & Hori, S. (2019). Effect of contrast injection protocol with appropriate iodine delivery rate and iodine concentration on hepatic enhancement at multi-detector row CT. European Radiology, 29(5), 2334–2342. https://doi.org/10.1007/s00330-018-5791-3
29. Singal, A. G., Llovet, J. M., Yarchoan, M., Mehta, N., Heimbach, J. K., Dawson, L. A., Jou, J. H., Kulkarni, V. M., Cabrera, R., Choi, J., Schelman, W. R., Kim, H. S., Venkatesh, S. K., Heiken, J. P., Roberts, L. R., & Zhang, Y. (2023). AASLD practice guidance on prevention, diagnosis, and treatment of hepatocellular carcinoma. Hepatology, 78(5), 1922–1965. https://doi.org/10.1097/HEP.0000000000000466
30. European Association for the Study of the Liver. (2018). EASL Clinical Practice Guidelines: Management of hepatocellular carcinoma. Journal of Hepatology, 69(1), 182–236. https://doi.org/10.1016/j.jhep.2018.03.019
31. Corwin, M. T., Fananapazir, G., Jin, M., & Lamba, R. (2016). Differences in liver imaging and reporting data system categorization between MRI and CT. Abdominal Radiology, 41(11), 2101–2107. https://doi.org/10.1007/s00261-016-0820-0
32. Kielar, A., Fowler, K. J., Lewis, S., Yee, J., & Chernyak, V. (2022). LI-RADS: A primer for beginners. Abdominal Radiology, 47(8), 2685–2703. https://doi.org/10.1007/s00261-022-03516-5
33. Brancatelli, G., Federle, M. P., Grazioli, L., Golfieri, R., & Lencioni, R. (2016). Benign regenerative nodules in Budd-Chiari syndrome and other vascular disorders of the liver: Radiologic-pathologic and clinical correlation. Radiographics, 22(4), 847–862. https://doi.org/10.1148/radiographics.22.4.g02jl16847
34. Schelhorn, J., Neumann, J., & Schmidtmann, I. (2023). CT and MRI of hepatocellular carcinoma: An update for radiologists. Journal of Magnetic Resonance Imaging, 57(2), 331–347. https://doi.org/10.1002/jmri.28421
35. Choi, J. Y., Lee, J. M., Sirlin, C. B., & Kim, S. H. (2019). CT and MR imaging diagnosis and staging of hepatocellular carcinoma: Part II. Extracellular agents, hepatobiliary agents, and ancillary imaging features. Radiology, 273(1), 30–50. https://doi.org/10.1148/radiol.2015141530