1

Universal Liver Imaging Lexicon: Imaging Atlas for

Research and Clinical Practice
Kathryn J. Fowler, MD • Mustafa R. Bashir, MD • David T. Fetzer, MD • Azusa Kitao, MD, PhD • Jeong Min Lee, MD • Hanyu Jiang, MD
Ania Z. Kielar, MD • Maxime Ronot, MD • Aya Kamaya, MD • Robert M. Marks, MD • Khaled M. Elsayes, MD • An Tang, MD, MSc
Claude B. Sirlin, MD • Victoria Chernyak, MD, MS
Author affiliations, funding, and conflicts of interest are listed at the end of this article.

The use of standardized terms in assessing and reporting disease processes has well-established benefits, such as clear communica-
tion between radiologists and other health care providers, improved diagnostic accuracy and reproducibility, and the enhancement
and facilitation of research. Recently, the Liver Imaging Reporting and Data System (LI-RADS) Steering Committee released a uni-
versal liver imaging lexicon. The current version of the lexicon includes 81 vetted and precisely defined terms that are relevant to
acquisition of images using all major liver imaging modalities and contrast agents, as well as lesion- and organ-level features. Most
terms in the lexicon are applicable to all patients undergoing imaging of the liver, and only a minority of the terms are strictly in-
tended to be used for patients with high risk factors for hepatocellular carcinoma. This pictorial atlas familiarizes readers with the
liver imaging lexicon and includes discussion of general concepts, providing sample definitions, schematics, and clinical examples
for a subset of the terms in the liver imaging lexicon. The authors discuss general, technical, and imaging feature terms used com-
monly in liver imaging, with the goal of illustrating their use for clinical and research applications.
Work of the U.S. Government published under an exclusive license with the RSNA.

Introduction precise vocabulary or lexicon is integral to creating standard-

The Liver Imaging Reporting and Data System (LI-RADS) is ized, structured, or templated reports. A standardized lexicon
a comprehensive system for standardizing the terminology, permits clear communication between radiologists and other
techniques, interpretation, reporting, and data collection on members of the health care team, facilitates clinical adoption
observations in patients who have or are at increased risk for of terms, improves diagnostic accuracy, and promotes agree-
hepatocellular carcinoma (HCC) (1). LI-RADS algorithms ment among radiologists (8). Standardized terms exist for
have been developed as a necessary response to an urgent many organs and disease processes, including prostate cancer,
need to standardize the imaging features, criteria, and treat- breast cancer, thyroid nodules, ovarian masses, and lung can-
ment of patients at risk for or with HCC. The benefits of stan- cer, but not—until recently—for liver imaging applicable for
dardized reporting are clear across the field of radiology, with the general population (6,8–20).
documented improvements in communication (2), data min- Beyond its use in clinical practice, a standardized, uni-
ing and auditing (3), adherence to clinical guidelines (4), and versally adopted lexicon is needed for research. Scientific
diagnostic accuracy (5–7). The development of a universal, studies in diagnostic radiology often test imaging features

An earlier incorrect version of this article appeared online. This article was corrected on December 6, 2022.

GASTROINTESTINAL IMAGING
This copy is for personal
use only. To order copies,
contact reprints@rsna.org
January 2023 Fowler et al 2

RadioGraphics 2023; 43(1):e220066 Developing a standardized lexicon for all of liver imaging is
https://doi.org/10.1148/rg.220066 a critical step toward harnessing the future potential of “‘big
Supplemental data” comprising our imaging studies and for developing ac-
Material Content Code: GI
Abbreviations: AP = arterial phase, APHE = AP hyper- curate population-based models for diagnosis and prognosis.
enhancement, CEUS = contrast-enhanced US, COU =
context of use, DP = delayed phase, HBP = hepatobili-
In an effort to standardize communication in the clinical set-
ary phase, HCC = hepatocellular carcinoma, LI-RADS = ting and the scientific literature, the LI-RADS Steering Com-
Liver Imaging Reporting and Data System, LP = late mittee recently released a liver imaging lexicon (24). The cur-
phase, PVP = portal venous phase, TP = transitional
phase rent version of the lexicon comprises 81 vetted and precisely
defined terms relevant to imaging with all major liver imaging
TEACHING POINTS modalities and contrast agents and includes lesion- and or-
„ A standardized lexicon permits clear communication between radiologists and gan-level features. Most terms are applicable to all patients
other members of the health care team, facilitates clinical adoption of terms, undergoing liver imaging, although a minority of terms are
improves diagnostic accuracy, and promotes agreement among radiologists. intended for imaging of patients at a high risk for HCC. The
„ COU refers to the condition under which a given term should be used in clini- lexicon is available free of charge on the American College of
cal practice or research. Terms with a broad COU are applicable for all patients
Radiology (ACR) website (https://www.acr.org/-/media/ACR/
undergoing liver imaging, while those with a LI-RADS–specific COU are appli-
cable mainly for patients with or at high risk for HCC. Files/RADS/LI-RADS/LIRADS-Lexicon-Table.pdf) (24).
„ Washout can be assessed only during an extracellular phase after the AP,
The purpose of this article is to provide a pictorial atlas
which is the PVP or the DP at CT, extracellular contrast−enhanced MRI, and of this universal liver imaging lexicon. We aim to familiar-
gadobenate-enhanced MRI and only the PVP at gadoxetate-enhanced MRI. ize readers with the lexicon by discussing general concepts,
„ Growth applies only to masses and does not apply to pseudolesions (ie, alter- reviewing sample definitions, and providing schematics and
ations in perfusion) or nonmasslike lesions (ie, focal fat deposition). clinical examples to illustrate the use of terms for clinical and
„ The presence of a targetoid appearance suggests intrahepatic cholan- research applications.
giocarcinoma or other non-HCC malignancies, including combined HCC-
cholangiocarcinoma and metastases to the liver from an extrahepatic primary Lexicon Development
tumor, but it does not exclude HCC.
Multiple terms and definitions have been developed and re-
fined since the LI-RADS inception in 2008 and have been pub-
lished in each released version of LI-RADS. The terms and
against a reference standard or outcome. While some terms their definitions were revised as evidence and user feedback
for imaging features might seem clear, their precise defi- accumulated. For example, the term rim APHE was formally
nitions and use can be ambiguous. For example, the com- introduced in the 2017 version, when the scientific evidence for
monly used term arterial enhancement can be applied and the role of APHE morphology in the differentiation of intrahe-
interpreted inconsistently in different studies. In some, the patic cholangiocarcinoma from HCC became available (22,23).
term may apply only to lesions with enhancement greater Clear definition of imaging features of non-HCC malignancy
than that of the liver during the arterial phase (AP), while from those of HCC allows high specificity in distinguish-
in other studies, the term may also include lesions with en- ing between the conditions and for achieving a 95% positive
hancement similar to that of the liver in the AP. Additional predictive value for diagnosis of LI-RADS-5 (definite HCC)
important attributes such as the morphology of the enhanc- (25,26).
ing component (eg, rim vs nonrim enhancement) and tim- In 2019, the process of development and approval of the
ing of the AP (eg, early vs late phase), if not specified, may lexicon was formalized, and the Lexicon Working Group
affect the implications of the term for the final diagnosis. For was convened. The members of the Lexicon Working Group
example, both HCC and intrahepatic cholangiocarcinoma extracted relevant terms and their definitions from the ex-
exhibit AP hyperenhancement (APHE), yet peripheral dis- isting LI-RADS documents and relevant literature. After
tribution of APHE in a lesion (ie, rim APHE) is typical of the Lexicon Working Group created and reviewed multi-
intrahepatic cholangiocarcinoma, whereas APHE that is not ple iterations of the lexicon terms and their definitions, the
confined to the lesion’s periphery (ie, nonrim APHE) is typ- entire 34-member LI-RADS Steering Committee reviewed
ical of HCC (21–23). Therefore, specifying the morphology and eventually approved the lexicon, with each term and its
of enhancement is necessary to distinguish between these definition requiring greater than 90% approval. Ultimately,
two neoplasms (21–23). Consequently, a scientific study per- 81 terms were approved and included in the most recent
formed to assess the presence of APHE for diagnosis of HCC version of the lexicon, which was released in June 2021.
without assessment of morphology would likely result in dif-
ferent diagnostic performance than a study that accounted Lexicon Structure
for both the presence and the morphology of APHE. Such in- For each included term, the LI-RADS lexicon provides the
consistent terminology challenges the synthesis of the pub- term name, definition, context of use (COU), applicable mo-
lished literature, introducing noise and potential errors into dalities, general comments, LI-RADS–specific comments,
the synthesized data. Furthermore, a lack of a standardized synonyms, and the type of term. COU refers to the condition
lexicon in published literature could have a negative effect under which a given term should be used in clinical prac-
on subsequent research studies, if the previously identified tice or research. Terms with a broad COU are applicable for
variables are not described consistently or reproducibly. all patients undergoing liver imaging, while those with a

Volume 43 Number 1 • radiographics.rsna.org

January 2023 Fowler et al 3

LI-RADS–specific COU are applicable mainly for patients Imaging Phases

with or at high risk for HCC. Of the 81 terms included in the
latest version, 69 have a broad COU, and 12 have a LI-RADS– Arterial Phase.—The AP is the contrast-enhanced phase when
specific COU. Although terms with a broad COU may be used the hepatic artery and branches are fully enhancing and the
for all patients undergoing liver imaging, the implications hepatic veins are not enhancing more than the liver by an-
may differ for patients with and without a risk for HCC. For tegrade flow. At CEUS, the AP usually starts approximately
instance, the feature “fat in mass, more than adjacent liver” 10–15 seconds after injection of contrast material and lasts for
favors the diagnosis of HCC when it is present in a patient at 10–20 seconds. At CT and MRI, the AP is divided into two
high risk. However, the same feature in a young woman with- temporal subtypes: early AP, when the portal vein is not en-
out background liver disease is more suggestive of a hepato- hancing or is enhancing less than the liver, and late AP, when
cellular adenoma. To keep the definitions short, the lexicon the portal vein is enhancing more than the liver. The AP is es-
includes comments with clarifying information that can be sential for characterization of lesions with preferential arterial
general or LI-RADS–specific. blood supply (27,28). In general, the late AP is preferable to
The lexicon is subdivided into broad categories of terms the early AP because the former demonstrates nonrim APHE,
including general, technical, and imaging feature terms. which is an imaging feature that is required for noninvasive
The following sections provide a sample of terms from each diagnosis of HCC, better than the early AP (1,29) (Figs 2–4).
broad category. We present the terms and their definitions
as they appear in the lexicon. Portal Venous Phase.—The PVP is the phase after the AP
that occurs no more than 2 minutes after injection of a con-
General Terms trast agent, when the portal and hepatic veins are enhancing
General terms are those that refer to focal abnormalities. more than the liver. At CEUS, the PVP usually starts approx-
They have a broad COU and are applicable to all modal- imately 30–45 seconds after injection, lasts for 90–100 sec-
ities (ie, CT, MRI, gray-scale US, and contrast-enhanced onds, and ends at approximately 2 minutes after injection.
US [CEUS]). The lexicon standardized these terms and es- At CT and MRI, the PVP images are acquired approximately
tablished a hierarchical relationship (Fig 1). For example, 60–80 seconds after start of injection.
observation is defined as an area that is distinctive when
compared with the background liver at imaging. Observa- Late Phase.—The LP is the phase of CEUS after the AP and
tions may be true lesions (ie, if it corresponds to a patho- the PVP when the portal and hepatic veins are enhancing
logic abnormality) or pseudolesions (if it does not corre- but less so than they are during the PVP. The LP lasts from
spond to a pathologic abnormality). If there is uncertainty the end of the PVP until there is clearance of microbubbles
about whether an observation represents a true lesion or from the circulation at approximately 4–6 minutes. In the
a pseudolesion, the term observation is preferred over the LP, the liver parenchyma is enhancing, but usually less than
term lesion. Lesions are further subdivided into nonmass- it does in the PVP.
like lesions (ie, those not distorting or destroying the paren-
chyma or other anatomic structures) and masses (ie, those Delayed Phase.—The DP is the phase after the AP that occurs
distorting or destroying the parenchyma or other anatomic at least 2 minutes after injection of an extracellular agent or
structures). Ovoid or rounded noncystic lesions measur- gadobenate dimeglumine, when the portal and hepatic veins
ing 2 cm or smaller can be referred to as nodules. While are enhancing more than the liver. The DP applies to CT,
pseudolesions can be masslike (ie, hypertrophic pseudo- extracellular contrast–enhanced MRI, and MRI with gado-
mass) or nonmasslike (ie, perfusional anomalies), there are benate dimeglumine enhancement. DP images are typically
no terms applicable to such subgroups. acquired 2–5 minutes after injection of contrast material. The
term DP does not apply to gadoxetate-enhanced MRI.
Technical Terms
Technical terms are those that refer to imaging, and they have Transitional Phase.—The TP is the phase after the AP when im-
a broad COU. The lexicon contains technical terms for CT, ages are acquired with an intravenous hepatobiliary contrast
MRI, gray-scale US, and CEUS. Among the technical terms, agent, when the liver vessels and the hepatic parenchyma are
imaging phases are particularly important to define because of similar signal intensity, which occurs between the PVP and
they have profound implications for the detection, appear- the HBP. The term applies to MRI with gadoxetate; although
ance, and interpretation of liver observations. The next sec- the TP does occur when gadobenate is used, TP images are
tion includes definitions for the various phases and a few key usually not acquired with this agent. The TP typically occurs
clarifying comments from the lexicon. The AP and the portal 2–5 minutes after injection of gadoxetate, when the liver is
venous phase (PVP) are applicable to CT, MRI with extracel- enhancing because of both extracellular and intracellular dis-
lular contrast agents, MRI with hepatobiliary contrast agents, tribution of the contrast agent. The onset of the TP may occur
and CEUS. The transitional phase (TP) and hepatobiliary less than 2 minutes or more than 5 minutes after injection in
phase (HBP) are only applicable to MRI with hepatobiliary a minority of patients.
contrast agents, while the delayed phase (DP) is applicable
only to MRI with extracellular contrast agents and MRI with Hepatobiliary Phase.—The HBP is the contrast-enhanced
gadobenate. The late phase (LP) is applicable only to CEUS. phase in which a hepatobiliary agent is used, and the liver

Volume 43 Number 1 • radiographics.rsna.org

January 2023 Fowler et al 4

Figure 1. Flowchart shows

hierarchical relationships of
general terms referring to a
focal abnormality.

Figure 2. Early and late arterial phases (APs). Axial T1-weighted MR images after injection of an extracellular contrast agent show the early AP (A), where
hepatic arterial branches (white arrows) are fully enhancing and the hepatic vein (arrowhead) and portal vein (black arrow) are not enhancing; the late AP (B),
where hepatic arterial branches (white arrow) are fully enhancing, the hepatic vein (arrowhead) is not enhancing, and the portal vein (black arrow) is enhancing
more than the liver; and the portal venous phase (PVP) (C), where the portal vein (black arrow) and the hepatic vein (arrowhead) are enhancing more than the
liver. Note a 24-mm observation (*) that is not enhancing more than the liver during the early AP (A) and is enhancing more than the liver during the late AP
(B). In general, the late AP demonstrates nonrim AP hyperenhancement (APHE) better than the early AP (29).

parenchyma is intended to be hyperintense compared with MRI with gadoxetate, the recommendation is to acquire pre-
hepatic blood vessels. The HBP typically occurs approxi- contrast, AP, PVP, TP, and HBP images.
mately 20 minutes after injection of gadoxetate. If gadobenate
is used, the HBP occurs 1–3 hours after injection. Excretion of Terms for Imaging Features
contrast material into the biliary tree may or may not occur. The terms for imaging features can be subdivided into those
HBP images can be suboptimal in patients with severe liver that have a broad COU and those with a LI-RADS–specific
dysfunction (eg, cirrhosis or biliary obstruction), severe ste- COU. For instance, threshold growth is a term that is specific
atosis, or severe iron overload (Fig E1). to the LI-RADS and would not be applicable for use in a broad
COU. Although terms with a broad COU can be used in the
Multiphasic Imaging.—Multiphasic imaging is defined as im- general population or in patients at high risk for HCC, some
aging during two or more phases after injection of intrave- of these terms (eg, washout) have specific diagnostic impli-
nous contrast material. This term is applicable to CT, MRI, cations in the LI-RADS, because they are used in assigning
and CEUS. For diagnosis and staging of patients at risk for LI-RADS categories. Care should be taken to reserve LI-
HCC, the LI-RADS recommends acquisition during the AP, RADS–specific diagnostic implications to patients who meet
PVP, and LP at CEUS and during the AP, PVP, and DP at CT. the criteria for application of the LI-RADS (eg, those with
At MRI with an extracellular contrast agent, the LI-RADS rec- cirrhosis, chronic hepatitis B viral infection, or those with a
ommends obtaining precontrast, AP, PVP, and DP images; at personal history of HCC).
Volume 43 Number 1 • radiographics.rsna.org
January 2023 Fowler et al 5

Figure 3. Late AP, portal venous phase (PVP), and delayed phase (DP) on CT images. (A) Axial contrast-enhanced late AP CT image shows fully enhancing he-
patic arterial branches (short arrow), nonenhancing hepatic veins (arrowheads), and portal veins (long arrow) that are enhancing more than the liver. (B) Axial con-
trast-enhanced PVP CT image shows that the portal (arrow) and hepatic (arrowheads) veins are enhancing more than the liver. (C) Axial contrast-enhanced DP CT
image acquired at 3 minutes after injection of contrast material shows that the portal (arrow) and hepatic (arrowheads) veins remain more enhancing than the liver.

Figure 4. PVP, transitional phase (TP), and hepatobiliary phase (HBP) on MR images. (A) Axial T1-weighted gadoxetate-enhanced PVP MR image shows
the portal (arrow) and hepatic (arrowhead) veins enhancing more than the liver. (B) Axial T1-weighted TP image acquired 3 minutes after injection of gadox-
etate contrast agent shows the portal vein (arrow), the hepatic vein (arrowhead), and the hepatic parenchyma (*) to be of similar signal intensity. (C) Axial T1-
weighted HBP MR image acquired 20 minutes after injection of gadoxetate contrast material shows the hyperintense liver parenchyma (*) in comparison with
the portal vein (white arrow) and the hepatic vein (arrowhead). Note the excretion of gadoxetate in the biliary tree (black arrow).

images in imaging during the AP, whereas at CT, in a patient

who has not undergone treatment, APHE can usually be as-
sessed without acquiring precontrast images. At CEUS, assess-
ment of APHE requires continuous imaging during the AP.
APHE can be seen in the entire observation or only in
part of it, and APHE is considered to be present if any part of
the observation has APHE. Enhancement in the AP must be
higher than that of the liver; a change from hypoattenuation,
hypointensity or hypoechogenicity on precontrast images to
isoenhancement during the AP does not qualify as APHE.
At CT and MRI, APHE has two main subtypes based on
morphology: rim APHE and nonrim APHE. Rim APHE is
defined as a subtype of APHE that is mainly in the periphery
of the observation (Figs 5, 6), and nonrim APHE is defined
Figure 5. Rim APHE at CT. Axial contrast-enhanced late AP as a subtype of APHE that is not mainly in the periphery of
CT image shows a 32-mm observation (arrow) that is enhanc-
the observation (Figs 7, 8). At CEUS, APHE has four main
ing more than the liver. The APHE is seen mainly in the periph-
ery of the observation, confirming rim APHE. Biopsy results subtypes: rim APHE (Fig 9), nonrim APHE (Fig 10), spoke-
revealed HCC. wheel centrifugal APHE (Fig 11), and peripheral discontin-
uous nodular APHE (Fig 12).
The term APHE and its subtypes have a broad COU. In the
Arterial Phase Hyperenhancement context of the LI-RADS CT, MRI, and CEUS diagnostic algo-
APHE is defined as enhancement that is higher than that of the rithms, rim APHE indicates that the observation is probably
liver during the AP. The term applies to CEUS, CT, and MRI. At or definitely malignant and not HCC specific (ie, a LI-RADS
MRI, assessment of APHE requires acquisition of precontrast malignant [ie, LR-M] feature) and nonrim APHE is a major

Volume 43 Number 1 • radiographics.rsna.org

January 2023 Fowler et al 6

Figure 6. Rim APHE, peripheral washout, and

DP central enhancement at MRI. (A) Axial T1-
weighted late AP MR image after administration
of an extracellular contrast agent shows multiple
observations (arrows) with rim APHE higher than
that in the liver, seen mainly in the periphery of
the observation. (B) Axial DP MR image acquired
4 minutes after administration of contrast mate-
rial shows peripheral washout (ie, reduction in
enhancement from that during the AP, resulting
in hypoenhancement relative to the liver, mainly
in the periphery of the observation [arrows] )
and DP central enhancement (ie, the inner part
of observations is more enhancing than is the
periphery [arrowheads] ). Biopsy results revealed
multifocal intrahepatic cholangiocarcinoma.

Figure 7. Nonrim APHE on CT images. (A) Axial contrast-en-

hanced late AP CT image shows a 70-mm observation (arrow) with
enhancement higher than that of the liver. APHE is not mainly in
the periphery of the observation, confirming nonrim morphology.
The observation was pathologically diagnosed as HCC at surgical
resection. (B) Axial contrast-enhanced late AP CT image shows
a 24-mm observation (arrow) with enhancement higher than
that of the liver. APHE is not mainly seen in the periphery of the
observation, confirming nonrim morphology. Note that areas of
heterogeneity and lower levels of enhancement (arrowheads) can
be seen, because in this case, APHE is present in only part of the
observation. However, as long as the distribution of the enhancing
areas is not predominantly in the periphery, APHE is characterized
as nonrim enhancement. The observation was pathologically diag-
nosed as HCC at surgical resection.

pands outward in a radial spoke-wheel pattern (Fig 11). This

temporal subtype of APHE is suggestive of the diagnosis of
focal nodular hyperplasia. Peripheral discontinuous nodular
APHE is defined as areas of enhancement that during the AP
are initially round or globular in shape and distributed dis-
continuously along the periphery of a lesion and then rapidly
expand to fill the lesion in its entirely or nearly in its entirety
(Fig 12). This temporal subtype of APHE is diagnostic of non-
sclerosed hemangioma.

Washout
Washout is defined as a reduction in enhancement from an
earlier to a later phase, resulting in hypoenhancement rela-
tive to the liver (Fig 13). At CT or MRI, washout can manifest
Figure 8. Nonrim APHE at MRI. Axial T1-weighted late AP MR
image acquired after administration of an extracellular contrast
as a change from hyperenhancing to hypoenhancing (Fig 14)
agent shows multiple observations (arrows) with higher en- or from isoenhancing to hypoenhancing (Fig 15). At CEUS,
hancement than that of the liver. APHE is not mainly in the pe- washout manifests as one of the following: hyper- to hypoen-
riphery of the observation, confirming nonrim morphology. Bi- hancement (Fig E2), isoenhancement to hypoenhancement
opsy revealed a β-catenin–activated hepatocellular adenoma.
(Fig E3), or hypoenhancement to unequivocally more hy-
poenhancement (Fig 16).
feature of HCC. In combination with other major features, it Washout can be assessed only if images are obtained during
can be used to categorize an observation as definitely HCC (ie, at least two contrast-enhanced phases (eg, AP followed by one
LR-5). Nonrim APHE is a required feature for the diagnosis or more phases after the AP) so that the reduction in enhance-
of HCC, and observations without nonrim APHE cannot be ment over time can be assessed. Consequently, washout cannot
diagnosed as definitely HCC noninvasively (1). be assessed in an examination with a single contrast-enhanced
Two of the APHE subtypes are assessable most reliably phase. Washout can be assessed only during an extracellular
with continuous imaging during the AP at CEUS. Spoke- phase after the AP, which is the PVP or the DP at CT, extracel-
wheel centrifugal APHE is defined as lesion enhancement that lular contrast–enhanced MRI, and gadobenate-enhanced MRI
during the AP begins as an internal focus and then rapidly ex- and only the PVP at gadoxetate-enhanced MRI. TP and HBP

Volume 43 Number 1 • radiographics.rsna.org

January 2023 Fowler et al 7

Figure 9. Rim APHE with early and marked washout on contrast-enhanced US (CEUS) images. (A) CEUS and gray-scale US images acquired 19
seconds after intravenous administration of 1.0 mL of sulfur hexafluoride lipid–type A microspheres show multiple observations (arrows) with higher
enhancement than that of the liver. APHE is seen mainly in the periphery of the observation, confirming rim morphology. (B) CEUS and gray-scale
images acquired 59 seconds after injection of contrast material show a reduction in the enhancement of the observations (arrows) from the earlier
phase (A) to hypoenhancement in the later phase, confirming washout. Washout was early (<60 seconds) and marked (ie, virtually devoid of en-
hancement or a “punched-out” appearance). Biopsy results revealed breast cancer metastases.

At CEUS, washout is divided into subtypes on the basis of

the time of onset and degree. A time of onset of less than 60
seconds after injection of contrast material is considered early
washout, and a time of onset of 60 seconds or longer after
injection is considered late washout. The degree of washout
can be mild or marked. Mild washout is less enhancing than
the liver but is not devoid of enhancement (ie, some enhance-
ment persists) (Fig 16). If mild washout becomes marked
more than 2 minutes after injection of the contrast agent, it is
still characterized as mild. Marked enhancement is virtually
devoid of enhancement (ie, a “punched out” appearance) by 2
minutes after injection of the contrast agent (Fig 17).
The term washout and its subtypes have a broad COU. In
the context of the LI-RADS CT and MRI diagnostic algorithm,
peripheral washout is an LR-M feature and nonperipheral
washout is a major feature of HCC. In the context of the LI-
RADS CEUS diagnostic algorithm, early or marked washout
Figure 10. Nonrim APHE on CEUS images. CEUS and gray-scale US im-
ages acquired 6 seconds after intravenous administration of 1.0 mL of sulfur is an LR-M feature, whereas late and mild washout is a major
hexafluoride lipid–type A microspheres show a 16-mm observation (arrow) feature of HCC.
with higher enhancement than that of the liver. APHE is not mainly in the
periphery of the observation, confirming nonrim morphology. The nodule
was presumed to be an adenoma, and the patient was advised to undergo
Capsule
biannual surveillance. A capsule is defined as a smooth, uniform, sharp border at CT
or MRI that encloses most or all of an observation (Figs 18, E4).
This term has a LI-RADS–specific COU. Capsules are of one of
hypointensity (see later sections in this article) do not qualify as two subtypes: an enhancing capsule, which is a major feature
washout. The presence of some enhancement in earlier phases of HCC, and a nonenhancing capsule, which is an ancillary
is required, as washout does not apply to nonenhancing obser- feature that favors diagnosis of HCC. An enhancing capsule is
vations. Washout can be seen in the entire observation or in visible as an enhancing rim during the PVP, DP, or TP, whereas
only parts of the observation and is characterized as present if a nonenhancing capsule does not show enhancement during
any part of the observation has washout. Finally, reduction in any postcontrast phases (Fig E5). If a capsule is visible as an
enhancement from APHE to isoenhancement does not qualify enhancing rim during one phase and as a nonenhancing rim
as washout (this temporal pattern is referred to as “fade”). during a different phase, it should be characterized as an en-
At CT or MRI, washout has two subtypes based on mor- hancing capsule. The imaging feature capsule refers to the im-
phology: peripheral washout (washout is mainly in the obser- aging appearance. The feature may correspond to a true tumor
vation periphery) (Fig 6) and nonperipheral washout (washout capsule or a pseudocapsule, and the distinction between the
that is not mainly in the observation periphery) (Figs 14, 15). two can be made only at pathologic examination. A capsule
Volume 43 Number 1 • radiographics.rsna.org
January 2023 Fowler et al 8

Figure 11. CEUS images show spoke-wheel centrifugal APHE that begins as an internal focus in the lesion at 6 seconds (arrow) and then rapidly expands
outward in a radial spoke-wheel pattern (arrowheads). This enhancement pattern allowed confirmation of the diagnosis of focal nodular hyperplasia.

may require the radiologist to make a judgment call during

interpretation.
In the context of the LI-RADS CT and MRI diagnostic al-
gorithm, growth is subdivided into threshold growth and sub-
threshold growth. Threshold growth is a major feature of HCC
and involves an increase in the size of the mass by 50% or
more in less than or equal to 6 months. Subthreshold growth
is an ancillary feature that, in general, favors diagnosis of a
malignancy, with a size increase of less than that of threshold
growth. Subthreshold growth is defined as a size increase of
less than 50% over any time period, any size increase over a
time interval of greater than 6 months, or a new mass of any
size (Fig 19). Threshold growth and subthreshold growth both
have a LI-RADS-specific COU and are applicable to CT and
MRI. These terms are not applicable to US or CEUS LI-RADS.

Targetoid Appearance
Figure 12. Peripheral discontinuous nodular APHE at CEUS. CEUS
images acquired at 3 and 13 seconds after intravenous administration Targetoid is defined as a target-like appearance at CT or MRI,
of 1.0 mL of sulfur hexafluoride lipid–type A microspheres show areas where the center and periphery of a mass have different im-
of enhancement (arrows) that are globular and distributed discontin- aging characteristics. The term has a broad COU. The pres-
uously along the periphery of the lesion on the 3-second image and ence of a targetoid appearance suggests intrahepatic chol-
rapidly expand to fill the lesion at 13 seconds. This pattern is diagnos-
tic for a nonsclerosing hemangioma. angiocarcinoma or other non-HCC malignancies, including
combined HCC-cholangiocarcinoma and metastases to the
liver from an extrahepatic primary tumor, but it does not
may appear as having either a complete or incomplete rim; exclude HCC. There are five subtypes of targetoid morphol-
the latter, known as capsule disruption, is associated with a ogy, including rim APHE and peripheral washout, which
worse prognosis in patients with HCC (30,31). were described in previous sections of this article. The third
subtype is delayed central enhancement (Fig 6), which is a
Growth post-AP pattern in which the inner part of the observation
Growth is defined as a definite increase in the size of a mass is more enhancing than is the periphery. This feature does
that cannot be explained by differences in techniques, arti- not apply to a central scar with delayed enhancement or to
facts, measurement errors, or interval hemorrhage. When- observations that can be confidently diagnosed as heman-
ever possible, the mass should be measured on images from giomas on the basis of other features. The fourth subtype
the same phase, sequence, and plane in serial examina- is targetoid diffusion restriction (Fig 20), which appears the
tions. Growth applies only to masses and does not apply to greatest in the periphery of the observation periphery. The
pseudolesions (ie, alterations in perfusion) or nonmasslike fifth subtype is a targetoid appearance during the TP or HBP
lesions (ie, focal fat deposition). The term growth has a broad (Fig 21), in which the periphery of the observation is more
COU and is applicable to all modalities. Currently, there is hypointense than the center. In the context of LI-RADS pop-
insufficient evidence to define a specific threshold for estab- ulation, the presence of any of the targetoid features results
lishing the presence of growth. Therefore, the assessment in categorization as LR-M.

Volume 43 Number 1 • radiographics.rsna.org

January 2023 Fowler et al 9

Figure 13. Illustration shows patterns of enhancement that qualify as washout (A) and as fade (B).

Figure 14. Washout on CT images. Axial

CT images in the late AP (A) and PVP (B)
show an observation (arrow), with reduction
in enhancement from the earlier (A) to the
later (B) phase, resulting in hypoenhance-
ment relative to the liver. In this case, the
observation showed a change from hype-
renhancing during the AP (A) to hypoen-
hancing during the PVP (B). This observa-
tion was categorized as LR-5 (definite HCC).

Figure 15. Washout on MR images. Axial

T1-weighted MR images in the late AP (A)
and PVP (B) after injection of an extracel-
lular contrast agent show an observation
(arrow), with reduction in enhancement
from the earlier (A) to the later (B) phase,
resulting in hypoenhancement relative
to the liver. In this case, the observation
demonstrated a change from isoenhancing
during the AP (A) to hypoenhancing during
the PVP (B). The observation was confirmed
as HCC at resection.

Mild to Moderate T2 Hyperintensity lower than that of simple fluid. The term has a broad COU
Mild to moderate T2 hyperintensity is defined as signal inten- and is applicable to MRI. In the context of the LI-RADS CT
sity on T2-weighted MR images that is higher than that of the and MRI diagnostic algorithm, mild to moderate T2 hyperin-
liver and similar to or lower than that of the non–iron-over- tensity is an ancillary feature that favors a diagnosis of malig-
loaded spleen (Fig 22). In patients without a spleen and those nancy in general.
with an iron-overloaded spleen, the signal intensity should be
Volume 43 Number 1 • radiographics.rsna.org
January 2023 Fowler et al 10

Figure 16. Mild washout on CEUS images. CEUS and gray-scale US images acquired 11 seconds (A) and 35 seconds (B) after intravenous adminis-
tration of 1.0 mL of sulfur hexafluoride lipid–type A microspheres show a 17-mm nodule (arrow) in the right hepatic lobe, with a reduction in enhance-
ment relative to that of the liver from hypoenhancement in an earlier phase (A) to unequivocally more hypoenhancement in the later phase (B). Note
that the nodule is less enhancing than the liver, but not devoid of enhancement, which indicates mild washout. The results of a biopsy of the nodule
revealed extramedullary hematopoiesis.

Figure 18. Enhancing capsule at MRI. Axial T1-weighted

MR image acquired during the PVP after administration of an
extracellular contrast agent shows a smooth, uniform, sharp
Figure 17. Marked washout at CEUS. CEUS and gray-scale US images border (arrow) around an observation (*), which is visible as an
acquired 2 minutes after the intravenous administration of 1.0 mL of sulfur enhancing rim. On the basis of the combination of all major fea-
hexafluoride lipid–type A microspheres show a 27-mm nodule (arrow) that tures (not all shown), this observation met the criteria for LR-5
is virtually devoid of enhancement (punched-out appearance) in the right (definite HCC).
hepatic lobe. The nodule was diagnosed as an intrahepatic cholangiocarci-
noma by means of pathologic examination at surgical resection.
can help to differentiate between diffusion restriction and
T2-shine-through: the ADC is similar to or lower than that of
Diffusion Restriction the liver when diffusion restriction is present, and it is higher
Diffusion restriction is defined as signal intensity higher than than that of the liver when T2 shine-through is present.
that of the liver on diffusion-weighted MR images that is In the context of the LI-RADS CT and MRI diagnostic
not caused only by T2 shine-through. In cases where an ad- algorithm, diffusion restriction has three subtypes: marked
equate apparent diffusion coefficient (ADC) map is obtained diffusion restriction (Fig 24), which is a nontargetoid LR-M
or if the ADC is calculated from source images, the ADC feature; targetoid diffusion restriction (Fig 20), which is a
is lower than or similar to that of the liver. The term has a targetoid LR-M feature; and neither marked nor targetoid
broad COU and is applicable to MRI. Diffusion restriction diffusion restriction (Fig 25), which is an ancillary feature

moderate diffusion weighting (b ≥400 sec/mm2).

should be assessed on MR images acquired with at least favoring diagnosis of malignancy, in general.

T2 shine-through can be seen in observations with mod- TP Hypointensity

erate to high signal intensity on T2-weighted MR images (Fig TP hypointensity is defined as signal intensity during the TP that
23). The calculated ADC or the appearance of the ADC map is lower than that of the liver. The term has a broad COU and is

Volume 43 Number 1 • radiographics.rsna.org

January 2023 Fowler et al 11

Figure 19. Illustration shows the criteria for

threshold growth and subthreshold growth.

Figure 20. Targetoid diffusion restriction. Axial diffusion- Figure 22. Mild to moderate T2 hyperintensity. Axial T2-
weighted MR image (b = 800 sec/mm2) shows an observation weighted MR image shows an observation (arrow) with a signal
with signal intensity higher than that of the liver, with the high- intensity higher than that of the liver and similar to or lower
est signal intensity in the periphery of the observation (arrow). than that of the non–iron-overloaded spleen (*). Note that the
Biopsy results revealed metastasis of a neuroendocrine tumor. observation is not as hyperintense as the cerebrospinal fluid
(arrowhead). Biopsy results revealed breast cancer metastasis.

applicable to gadoxetate-enhanced MRI (Fig 26). As we stated

previously, TP hypointensity does not qualify as washout. In
the context of the LI-RADS CT and MRI diagnostic algorithm,
TP hypointensity is an ancillary feature favoring diagnosis of
malignancy, in general, unless it is in a targetoid pattern.

HBP Hypointensity
HBP hypointensity is defined as signal intensity during the
HBP that is lower than that of the liver (Fig 26). The term has
a broad COU and is applicable to gadoxetate-enhanced MRI
and gadobenate-enhanced MRI. As with TP hypointensity,
Figure 21. Targetoid HBP hypointensity. Axial T1-weighted HBP HBP hypointensity does not qualify as washout. In the con-
MR image acquired 20 minutes after administration of gadox- text of the LI-RADS CT and MRI diagnostic algorithm, HBP
etate shows an observation (arrow) that is more hypointense in hypointensity is an ancillary feature favoring diagnosis of ma-
the periphery than in its center. The observation was diagnosed
as intrahepatic cholangiocarcinoma at surgical resection.
lignancy, in general, unless it is in a targetoid pattern.

Volume 43 Number 1 • radiographics.rsna.org

January 2023 Fowler et al 12

Figure 23. T2 shine-through. (A) Axial T2-weighted MR image with fat suppression shows a 43-mm observation (arrow) with markedly high signal intensity,
which is similar to that of the cerebrospinal fluid (arrowhead). (B) Diffusion-weighted MR image (b = 800 sec/mm2) shows the signal intensity in the observation
(arrow) to be higher than that in the liver (*). (C) Apparent diffusion coefficient (ADC) map shows the observation (arrow) to have higher signal intensity than
that of the liver (*), which confirms the presence of T2 shine-through rather than diffusion restriction. The observation was diagnosed at surgical resection as a
moderately differentiated HCC, with a pseudoglandular pattern.

Peripheral Discontinuous Nodular Enhancement

Peripheral discontinuous nodular enhancement is defined as ar-
eas of enhancement that are round or globular in shape and dis-
tributed discontinuously along the periphery of a lesion in the
early postcontrast phases and expand and appear approximately
the same as the blood pool in intensity in subsequent phases
(Fig 27). The term has a broad COU; is applicable to CEUS, CT,
and MRI; and is a diagnostic imaging feature of nonsclerosing
hemangiomas. This peripheral discontinuous nodular enhance-
ment feature reflects a temporal enhancement pattern; there-
fore, its strict assessment requires acquisition of images during
two or more phases. However, a diagnosis of hemangioma can
be made on the basis of imaging during a single postcontrast
phase if the imaging features are sufficiently characteristic. In
Figure 24. Marked diffusion restriction. Axial diffusion-weighted such cases, the temporal pattern is inferred.
MR image (b = 800 sec/mm2) shows an observation (arrow) with
signal intensity higher than that of the liver and unequivocally The enhancing areas approximate the appearance of the
higher than that of the non–iron-overloaded spleen (*). The blood pool in intensity on images from all postcontrast phases.
observation was diagnosed as HCC by means of pathologic If a hepatobiliary agent is administered, the enhancing areas
evaluation at surgical resection. usually become isointense and then hypointense relative to
the liver in the TP and HBP, similar to the appearance of the
vessels. As the areas of enhancement expand, they may co-
alesce to become continuous or may fill the lesion entirely,
and they may no longer appear round or globular.

Fade
Fade is defined as a reduction in enhancement relative to that
of the liver from hyperenhancement in an earlier phase to
isoenhancement or minimal hyperenhancement in all later
phases. The term has a broad COU and is applicable to CEUS,
CT, and MRI (Figs 28–30). Fade can be assessed only if images
are obtained during at least two contrast-enhanced phases
(eg, the AP followed by one or more phases after the AP), so
that the reduction in enhancement over time can be assessed.
Fade can appear as one of two patterns: (a) hyperen-
Figure 25. Diffusion restriction. Axial diffusion-weighted hancement during the AP and isoenhancement to minimal
image (b = 800 sec/mm2) shows an observation (arrow) with hyperenhancement during all later phases or (b) hyperen-
signal intensity higher than that of the liver and less than that hancement during the PVP and isoenhancement to minimal
of the non–iron-overloaded spleen (*). The observation was
diagnosed as HCC by means of pathologic examination at
hyperenhancement during all later phases (Fig 13).
surgical resection. When hypoenhancement relative to the liver is seen
during any phase after the AP, the pattern should not be
characterized as fade.
Volume 43 Number 1 • radiographics.rsna.org
January 2023 Fowler et al 13

Figure 26. TP and HBP hypointensity. Ax-

ial T1-weighted MR images acquired during
the TP (A) and HBP (B) show multiple
observations (arrows) with signal intensity
lower than that of the liver. Biopsy results
revealed metastases of a neuroendocrine
tumor.

Figure 27. Peripheral discontinuous nod-

ular enhancement on MR images. Axial
T1-weighted MR images in the late AP (A)
and PVP (B) after administration of extracel-
lular contrast agent show globular areas of
enhancement (arrows) that are distributed
discontinuously along the periphery of the
lesion during the AP (A) and expanded en-
hancement during the PVP (B) that is approx-
imately the same as that of the blood pool
(arrowheads). The enhancing components fill
the lesion nearly entirely during the DP (Fig
E6). The enhancement pattern is diagnostic
of a nonsclerosing hemangioma.

Figure 28. Fade on CT images. Axial CT

images in the late AP (A) and PVP (B) show
an observation (arrow) with reduction in
enhancement relative to the liver from hy-
perenhancement in the earlier phase (A) to
isoenhancement in all later phases, includ-
ing the PVP (B) and the DP (Fig E7). The ob-
servation was diagnosed as a moderately
differentiated HCC by means of pathologic
examination at surgical resection.

Although washout and fade are similar in that the area of adoption of the standardized lexicon is crucial to promoting
interest appears to de-enhance relative to earlier phases, the clarity and consistency of communication in clinical practice
two terms are not the same (Fig 13). Washout follows isoen- and scientific literature (8). Over time, the lexicon will be im-
hancement or hyperenhancement during an earlier phase proved, and additional terms will be added as the evidence
(at CT or MRI) or any degree of enhancement in an earlier accumulates, expert consensus matures, and user feedback
phase (at CEUS) and results in hypoenhancement relative to is compiled.
the liver in a later phase. In comparison, fade follows hyper­ The current lexicon focuses on subjective imaging features
enhancement in an earlier phase (with all modalities), and and descriptions. As data on quantitative imaging biomark-
results in isoenhancement or minimal hyperenhancement ers are compiled, terms related to quantitative images may
relative to the liver in all later phases. be added. Similarly, as evidence accrues and matures, the
lexicon may be expanded to incorporate prognostic imaging
Future Directions terms and their implications, such as such features that may
The latest version of the LI-RADS lexicon was released in help to predict long-term outcomes in patients with HCC.
June 2021. The authors of the lexicon and this article recog-
nize that a perfect consensus by all users for every term in Conclusion
the lexicon is not possible, and some readers may dislike or The LI-RADS universal liver imaging lexicon is available free
disagree with a particular term or definition. Nevertheless, of charge to members and nonmembers on the American
Volume 43 Number 1 • radiographics.rsna.org
January 2023 Fowler et al 14

Figure 29. Fade on CEUS images. CEUS and gray-scale US images acquired at 12 seconds (A) and at 4 minutes and 22 seconds (B) after intrave-
nous administration of 1.0 mL of sulfur hexafluoride lipid–type A microspheres show a 26-mm nodule (arrows) in the left hepatic lobe, with reduction
in enhancement relative to that of the liver from hyperenhancement in an earlier phase (A) to isoenhancement in the later phase (B). The nodule
was presumed to be an adenoma and the patient was advised to undergo biannual surveillance.

Figure 30. Fade on MR images. Axial

T1-weighted MR images in the late AP (A)
and PVP (B) after injection of an extracel-
lular contrast agent show an observation
(arrow), with reduction in enhancement
relative to that of the liver from hyperen-
hancement during the earlier phase (A)
to isoenhancement during later phases
including the PVP (B) and the DP (Fig E8).
The observation showed concentration
of gadoxetate on a follow-up study (not
shown), which allowed confirmation of the
diagnosis of a focal nodular hyperplasia.

College of Radiology website. The lexicon is intended to pro- Canada (A.T.); and Department of Radiology, Memorial Sloan Kettering Can-
cer Center, 1275 York Avenue, New York, NY 10065 (V.C.). Received March
vide standardized, precise, and comprehensive terminology 24, 2022; revision requested May 11 and received May 27; accepted June 8.
for the field of liver imaging to be applied both in clinical care Address correspondence to V.C. (email: vichka17@hotmail.com).
and to facilitate standardization of research. Although most
terms originated in the context of the LI-RADS, many have H.J. is supported by the National Natural Science Foundation of China
(82101997) and the Science and Technology Department of Sichuan Province
broader applicability and may be used in all patients under- (2021YFS0141).
going liver imaging. Adoption of a universal liver imaging
lexicon will help to advance the care of patients and the syn- Disclosures of conflicts of interest.—K.J.F. Grants from GE, Median, and Pfizer;
thesis of evidence. consulting fees from Epigenomics; payment for lectures from Bayer, the Univer-
sity of California San Diego, New York University, and San Diego Radiological
Author affiliations.—From the Liver Imaging Group, Department of Radiol-
Society; leadership of the American College of Radiology Liver Imaging Report-
ogy, UC San Diego, San Diego, Calif (K.J.F., C.B.S.); Department of Radiol- ing and Data System Committee. M.R.B. Research support from Carmot Ther-
ogy, Duke University Health System, Durham, NC (M.R.B.); Department apeutics, Corcept Therapeutics, Diabetes and Endocrinology Consultants, Mad-
of Radiology, University of Texas Southwestern Medical Center, Dallas, Tex rigal Pharmaceuticals, Metacrine, NGM Biopharmaceuticals, Pinnacle Clinical
(D.T.F.); Department of Radiology, Kanazawa University Graduate School Research, ProSciento, and Siemens Healthineers. D.T.F. Grants, consulting fees,
of Medical Sciences, Kanazawa, Japan (A.Kitao); Department of Radiology, and receipt of equipment, materials, drugs, or other services from GE Healthcare
Seoul National University College of Medicine, Seoul, South Korea (J.M.L.); and Philips Healthcare; grants from Siemens Healthineers, grants (unrelated)
Department of Radiology, West China Hospital, Sichuan University, Chengdu, from the National Institutes of Health (5U01-DK061713-19, 1R01CA249765-01).
China (H.J.); Joint Department of Medical Imaging, University of Toronto, To- J.M.L. Grants from Bayer, Canon, Centers for Medicare and Medicaid Services,
ronto, Ontario, Canada (A.Z.K.); Department of Radiology, Université Paris Dongkuk, GE Healthcare, Guerbet, Philips Healthcare, RF Medical, Samsung
Cité, Paris, France, and Department of Radiology, Hôpital Beaujon, APHP. Medicine, and Starmed; consulting fees from Samsung Medicine; payment for
Nord, Clichy, France (M.R.); Department of Radiology, Stanford University lectures from Bayer, Guerbet, and Samsung Medicine. A.Z.K. Vice President
School of Medicine, Stanford, Calif (A.Kamaya); Department of Radiology, of the Canadian Board of Chancellors for the American College of Radiology.
Naval Medical Center San Diego, San Diego, Calif (R.M.M.); Department of A.K. Royalties from Elsevier. K.M.E. Member of the RadioGraphics editorial
Radiology, Uniformed Services University of the Health Sciences, Bethesda, board. A.T. Fonds de recherche du Québec en Santé and Fondation de l’Asso-
Md (R.M.M.); Department of Radiology, University of Texas MD Anderson ciation des radiologistes du Québec (FRQS-ARQ # 298509). C.B.S. Grants from
Cancer Center, Houston, Tex (K.M.E.); Department of Radiology, Radiation the American College of Radiology, Bayer, Foundation of the National Institutes
Oncology and Nuclear Medicine, Université de Montréal, Montréal, Quebec, of Health, GE Healthcare, Gilead, Pfizer, Philips, Siemens Healthineers; lab ser-

Volume 43 Number 1 • radiographics.rsna.org

January 2023 Fowler et al 15

vice agreements with Enanta, Gilead, ICON, Intercept, Nusirt, Shire, Synageva, MRI reporting-a preliminary study on 20 consecutive cases with newly
and Takeda; institutional consulting for Bristol-Myers-Squibb, Exact Sciences, diagnosed glioblastoma. BMC Med Imaging 2022;22(1):53.
IBM-Watson, Pfizer; personal consulting for Blade, Boehringer, Epigenomics; 15. Shinagare AB, Sadowski EA, Park H, et al. Ovarian cancer reporting lexicon
and Guerbet; royalties and/or honoraria from Medscape and Wolters Kluwer; for computed tomography (CT) and magnetic resonance (MR) imaging
stock options in Livivos; advisory board member for Quantix Bio; chief medical developed by the SAR Uterine and Ovarian Cancer Disease-Focused
officer for Livivos (unsalaried position with stock options). V.C. Consulting fees Panel and the ESUR Female Pelvic Imaging Working Group. Eur Radiol
from Bayer, chair of the American College of Radiology Liver Imaging Report- 2022;32(5):3220–3235.
ing and Data System Committee. 16. Baugnon KL. NI-RADS to Predict Residual or Recurrent Head and Neck
Squamous Cell Carcinoma. Neuroimaging Clin N Am 2022;32(1):1–18.
The views expressed in this article are those of the authors and do not nec- 17. Nougaret S, Rousset P, Gormly K, et al. Structured and shared MRI staging
essarily reflect the position or policy of the U.S. military or the United States lexicon and report of rectal cancer: A consensus proposal by the French
Government. Radiology Group (GRERCAR) and Surgical Group (GRECCAR) for rectal
cancer. Diagn Interv Imaging 2022;103(3):127–141.
18. Miyake KK, Kataoka M, Ishimori T, et al. A Proposed Dedicated Breast
References PET Lexicon: Standardization of Description and Reporting of Radiotracer
1. Chernyak V, Fowler KJ, Kamaya A, et al. Liver Imaging Reporting and Data Uptake in the Breast. Diagnostics (Basel) 2021;11(7):1267.
System (LI-RADS) Version 2018: Imaging of Hepatocellular Carcinoma in 19. Candelaria RP, Spak DA, Rauch GM, et al. BI-RADS Ultrasound Lexicon
At-Risk Patients. Radiology 2018;289(3):816–830. Descriptors and Stromal Tumor-Infiltrating Lymphocytes in Triple-Negative
2. Corwin MT, Lee AY, Fananapazir G, Loehfelm TW, Sarkar S, Sirlin Breast Cancer. Acad Radiol 2022;29 Suppl 1(Suppl 1):S35–S41.
CB. Nonstandardized Terminology to Describe Focal Liver Lesions in 20. Chhabra A, Gupta A, Thakur U, et al. Osseous Tumor Reporting and
Patients at Risk for Hepatocellular Carcinoma: Implications Regarding Data System-Multireader Validation Study. J Comput Assist Tomogr
Clinical Communication. AJR Am J Roentgenol 2018;210(1):85–90. 2021;45(4):571–585.
3. Yuan W, Sun J. Efficient Auditing of Standardized Reporting in Radiol- 21. Chen Y, Zhu Y, Chen K, et al. Differentiation between hepatocellular
ogy. J Am Coll Radiol 2019;16(1):6. carcinoma and intrahepatic cholangiocarcinoma using contrast-enhanced
4. Corwin MT, Arora A, Loehfelm TW, Fananapazir G, Campbell MJ. ultrasound: A systematic review and meta-analysis. Clin Hemorheol Mi-
Adherence to guidelines for hormonal evaluation in patients with in- crocirc 2021;79(2):293–309.
cidentally detected adrenal nodules: effects of radiology report wording 22. Iavarone M, Piscaglia F, Vavassori S, et al. Contrast enhanced CT-scan
and standardized reporting. Abdom Radiol (NY) 2020;45(9):2910–2915. to diagnose intrahepatic cholangiocarcinoma in patients with cirrhosis. J
5. Lin E, Powell DK, Kagetsu NJ. Efficacy of a checklist-style structured Hepatol 2013;58(6):1188–1193.
radiology reporting template in reducing resident misses on cervical spine 23. Tsunematsu S, Chuma M, Kamiyama T, et al. Intratumoral artery on con-
computed tomography examinations. J Digit Imaging 2014;27(5):588–593. trast-enhanced computed tomography imaging: differentiating intrahepatic
6. Nair AV, McInnes M, Jacob B, et al. Diagnostic accuracy and inter-ob- cholangiocarcinoma from poorly differentiated hepatocellular carcinoma.
server agreement with the CO-RADS lexicon for CT chest reporting in Abdom Imaging 2015;40(6):1492–1499.
COVID-19. Emerg Radiol 2021;28(6):1045–1054. 24. LI-RADS Lexicon Table. American College of Radiology. https://www.
7. Shaish H, Feltus W, Steinman J, Hecht E, Wenske S, Ahmed F. Impact acr.org/-/media/ACR/Files/RADS/LI-RADS/LIRADS-Lexicon-Table.pdf.
of a Structured Reporting Template on Adherence to Prostate Imaging Published 2021. Accessed October 3, 2022.
Reporting and Data System Version 2 and on the Diagnostic Performance 25. Zou X, Luo Y, Morelli JN, Hu X, Shen Y, Hu D. Differentiation of hepato-
of Prostate MRI for Clinically Significant Prostate Cancer. J Am Coll cellular carcinoma from intrahepatic cholangiocarcinoma and combined
Radiol 2018;15(5):749–754. hepatocellular-cholangiocarcinoma in high-risk patients matched to MR
8. Chernyak V, Tang A, Do RKG, et al. Liver imaging: it is time to adopt field strength: diagnostic performance of LI-RADS version 2018. Abdom
standardized terminology. Eur Radiol 2022;32(9):6291–6301. Radiol (NY) 2021;46(7):3168–3178.
9. Rao AA, Feneis J, Lalonde C, Ojeda-Fournier H. A Pictorial Re- 26. Lee S, Kim YY, Shin J, et al. CT and MRI Liver Imaging Reporting and Data
view of Changes in the BI-RADS Fifth Edition. RadioGraphics System Version 2018 for Hepatocellular Carcinoma: A Systematic Review
2016;36(3):623–639. With Meta-Analysis. J Am Coll Radiol 2020;17(10):1199–1206.
10. Razek AAKA, El-Diasty T, Elhendy A, Fahmy D, El-Adalany MA. Prostate 27. Mitsuzaki K, Yamashita Y, Ogata I, Nishiharu T, Urata J, Takahashi M.
Imaging Reporting and Data System (PI-RADS): What the radiologists Multiple-phase helical CT of the liver for detecting small hepatomas in
need to know? Clin Imaging 2021;79:183–200. patients with liver cirrhosis: contrast-injection protocol and optimal timing.
11. Özel A, Türkyılmaz Mut D, Ağrıdağ Üçpınar B, et al. Interobserver AJR Am J Roentgenol 1996;167(3):753–757.
Variability of Ultrasound Features Based on American College of 28. Hollett MD, Jeffrey RB Jr, Nino-Murcia M, Jorgensen MJ, Harris DP.
Radiology Thyroid Imaging Reporting and Data System Lexicon in Dual-phase helical CT of the liver: value of arterial phase scans in the
American College of Radiology Thyroid Imaging Reporting and Data detection of small (< or = 1.5 cm) malignant hepatic neoplasms. AJR Am
System System: A Single-Center Study With Radiologists and Radiology J Roentgenol 1995;164(4):879–884.
Residents. Ultrasound Q 2021;37(4):324–328. 29. Schima W, Hammerstingl R, Catalano C, et al. Quadruple-phase MDCT
12. Andreotti RF, Timmerman D, Strachowski LM, et al. O-RADS US Risk of the liver in patients with suspected hepatocellular carcinoma: effect of
Stratification and Management System: A Consensus Guideline from contrast material flow rate. AJR Am J Roentgenol 2006;186(6):1571–1579.
the ACR Ovarian-Adnexal Reporting and Data System Committee. 30. Lewis RH, Glazer ES, Bittenbinder DM, et al. Outcomes Following Resection
Radiology 2020;294(1):168–185. of Hepatocellular Carcinoma in the Absence of Cirrhosis. J Gastrointest
13. Carter BW, Lichtenberger JP 3rd, Wu CC, Munden RF. Screening for Lung Cancer 2019;50(4):808–815.
Cancer: Lexicon for Communicating With Health Care Providers. AJR Am 31. Jin YJ, Lee JW, Park SW, et al. Survival outcome of patients with spon-
J Roentgenol 2018;210(3):473–479. taneously ruptured hepatocellular carcinoma treated surgically or by
14. Huckhagel T, Stadelmann C, Abboud T, Riedel C. Determining the ap- transarterial embolization. World J Gastroenterol 2013;19(28):4537–4544.
plicability of the RSNA radiology lexicon (RadLex) in high-grade glioma

TM
This journal-based SA-CME activity has been approved for AMA PRA Category 1 Credit . See rsna.org/learning-center-rg.

Volume 43 Number 1 • radiographics.rsna.org

 1

Erratum
February 2023 • Volume 43 • Number 2

Originally published in:

https://doi.org/10.1148/rg.220066
Universal Liver Imaging Lexicon: Imaging Atlas for Research and Clinical Practice
Kathryn J. Fowler, Mustafa R. Bashir, David T. Fetzer, Azusa Kitao, Jeong Min Lee, MD, Hanyu Jiang, Ania Z. Kielar, Maxime Ronot,
Aya Kamaya, Robert M. Marks, Khaled M. Elsayes, An Tang, Claude B. Sirlin, Victoria Chernyak
Erratum in:
https://doi.org/10.1148/rg.239001
Page 4, Figure 2 legend: The second-to-last sentence should read as follows:
Note a 24-mm observation (*) that is not enhancing more than the liver during the early AP (A) and is enhancing more than the liver during
the late AP (B). In general, the late AP demonstrates nonrim AP hyperenhancement (APHE) better than the early AP (29).

©
ERRATUM This copy is for personal
use only. To order copies,
RSNA, 2023 contact reprints@rsna.org