A practical guide to consistent terminology, diagnostic certainty, measurement language, and recommendation statements in radiology reports.
Browse the other checklists, phrase sets, and systematic approaches in this topic.
Two radiologists may interpret the same images similarly, yet the words they use in their reports may be understood by clinicians as representing different levels of certainty. Expressions such as “suggestive of,” “consistent with,” “favored to represent,” “possible,” “suspicious for,” and “cannot be excluded” are common in daily reporting. The problem is that these expressions may not carry the same meaning in every institution or for every reader.
Terminology standardization does not mean converting every report into a single style. The aim is to use more consistent words for the same clinical concepts, make the level of uncertainty visible, and reduce the risk of misinterpretation.
The radiology report is not read only by radiologists. The referring physician, consulting teams, nursing and support staff, other radiologists, and increasingly patients may all access the report.
A nuance that is clear to the radiologist may not be interpreted in the same way by another reader. For example:
Standardized language aims to reduce the number of unanswered questions created by the report itself.
A strong report sentence should, whenever possible:
In the findings section, imaging observation and diagnostic interpretation should be balanced.
A 12 mm nodular opacity is present in the posterior segment of the right upper lobe.
This sentence provides location and size but offers limited information about morphology and likely significance.
A malignant nodule is present in the right upper lobe.
Unless the imaging features provide histopathologic certainty, this wording may be more definitive than the examination supports.
An irregular 12 mm solid nodule is present in the posterior segment of the right upper lobe. Its morphologic features are suspicious for primary lung malignancy.
The observation and interpretation are presented as separate layers. The degree of certainty is aligned with what imaging can reasonably establish.
There is no single universal certainty lexicon, but it may be useful for a department to define and use a limited set of expressions. The following is a simple educational framework. The categories should not be interpreted as fixed numerical probabilities.
Example expressions:
Use when the imaging features are highly specific and the clinical context is appropriate.
Example expressions:
Use when one diagnosis is clearly most likely but imaging alone does not provide certainty.
Example expressions:
Use when the findings are compatible with a diagnosis but meaningful alternatives remain.
Example expressions:
Use when the significance or nature of the finding cannot be established confidently.
Example expressions:
Use when the source of uncertainty is a specific technical limitation.
Instead of using “cannot be excluded” by itself, state what cannot be excluded and why.
Yes, but without unnecessary detail.
The following expressions may carry little meaningful difference when used for the same structure:
Selecting one or two preferred options for the same context within a department or personal phrase library can improve consistency.
A structure should be described as normal only when it has actually been evaluated. When assessment is technically limited, the limitation should be stated instead.
Vague:
No pathology.
Clearer:
No acute intracranial hemorrhage, significant mass effect, or midline shift.
Vague:
The lungs are normal.
Clearer:
No focal air-space opacity, pleural effusion, or pneumothorax.
Long negative lists are not necessary in every report. Negative statements should be selected according to the clinical question and purpose of the examination.
Grades such as “mild,” “moderate,” and “severe” are useful when they are based on defined criteria. When used without criteria, however, they may be open to personal interpretation.
The following approaches can improve standardization:
Example:
A moderate right pleural effusion is present, measuring up to 18 mm in maximal thickness.
Instead of using only “moderate,” this sentence provides a measurable reference. The measurement method should be appropriate for the modality and clinical context.
Location can be described through the following components:
Example:
A 14 × 11 mm hypodense lesion is present in hepatic segment VII, adjacent to the right hepatic vein.
Inconsistency in laterality or segment designation within the same report is a common editing error. Laterality deserves a separate final check, especially in reports created with speech recognition or copied templates.
A department may define a shared approach to the following:
Example:
A 9 × 7 mm simple cyst is present in the lower-pole cortex of the left kidney. It measured 8 × 7 mm on the examination dated March 12, 2025, without significant interval change.
When the word “stable” is used, the comparison date and measurement should be visible whenever possible.
Weak recommendation:
Clinical correlation is recommended.
Clearer recommendation:
If biliary obstruction remains clinically suspected, correlation with bilirubin and cholestatic enzyme levels is recommended.
Weak recommendation:
Further evaluation is recommended.
Clearer recommendation:
If clinically indicated, contrast-enhanced liver MRI may be considered for further characterization of the indeterminate lesion.
A useful recommendation should answer as many of the following questions as possible:
The radiologist’s authority and responsibility regarding recommendations should be considered together with departmental policies and current relevant guidelines.
Excessive repetition of phrases such as “is seen,” “is noted,” and “is visualized” can lengthen the report.
Instead of using one of these verbs in every sentence, use direct wording when possible:
A 12 mm gallstone is present within the gallbladder lumen.
“May possibly represent ...” “A suspicious appearance is suggested.”
Choose one clear expression of certainty.
Listing many diagnoses without explanation may not support clinical decision-making. State the most likely diagnosis first, then briefly describe meaningful alternatives and the features that support or argue against them.
An abbreviation that is clear within one institution may be ambiguous at an outside center or to a patient. Nonstandard or uncommon abbreviations should be written out at first use.
Collecting strong sentences during residency can be useful, but uncontrolled growth of the library may produce many similar expressions for the same concept.
A more organized method is to:
Example variant set:
This structure supports conscious selection instead of creating a new uncertainty phrase at random for every case.
Terminology standardization is not intended to eliminate the radiologist’s individual style. It aims to make the clinical message more predictably understood. Creating a limited but defined set of expressions for observation, certainty, comparison, and recommendations can strengthen report quality and the feedback process, particularly during residency.
The best standardized phrase is not the one that sounds most academic. It is the one that accurately represents the imaging finding, allows the reader to understand the intended meaning, and does not complicate clinical decisions with unnecessary ambiguity.
This content is educational and is not patient-specific medical advice. Every phrase, template, classification result, and recommendation must be verified and adapted by a physician using the complete examination, clinical context, current guidelines, and institutional protocol.
Add curated phrases and editable templates to your workflow to support a clearer shared reporting language.
