BI-RADS v2025 in Breast Ultrasound: Lexicon Changes

What ACR Published in 2025

The American College of Radiology announced the release of the BI-RADS v2025 Manual. For breast ultrasound diagnostics, this is important not as a cosmetic renaming but as an update to the standardized language that doctors use to describe findings, form conclusions, and convey clinical risk.

The ACR's public announcement does not provide a complete list of all new and excluded ultrasound terms. Therefore, when implementing, the official edition should not be replaced with a local retelling. The correct approach is to use the v2025 Manual as the legally and methodologically primary document, and the ACR's public message as a signal to update working templates.

What Remains Unchanged: BI-RADS Logic Remains Categorical

BI-RADS remains a system not only for description but also for final assessment. The lexicon is needed for reproducible description of features; the BI-RADS category is needed for risk communication and further tactics. If the protocol contains a rich description but the final category does not match the level of suspicion, the task of BI-RADS is not fulfilled.

Main Change for Ultrasound: Not Only Final Categories Need Updating

An error in implementing BI-RADS v2025 is to leave the old ultrasound template and only change the document header. Updating the lexicon means that formulations in sections like “lesion,” “non-lesion finding,” “calcifications,” “associated features,” “special cases,” and additional methods, if used in the department, must be checked.

Practically, this concerns dropdown lists in RIS/PACS, voice macros, screening ultrasound protocols, diagnostic ultrasound after mammography, and targeted ultrasound after MRI. Terms in these scenarios should be consistent: the same finding should not be named differently depending on the doctor or office.

Lexicon: Description Should Lead to Category

BI-RADS does not encourage free literary description when it is impossible to understand from the text why the final class was chosen. For ultrasound, the principle remains the same: first, the features of the finding are described with standardized descriptors, then an assessment is formed. If the features are benign, the category should not be artificially elevated. If the features are suspicious, a mild formulation in the text should not mask the need for BI-RADS 4 or 5.

When transitioning to v2025, it is useful to conduct an internal audit: take typical conclusions from recent months and check if it is possible to restore the type of finding, its key features, localization, dimensions, comparison, and final risk from the text alone without viewing the image. If not, the problem is not with the BI-RADS version but with the quality of the protocol.

Template Areas to Review First

• Type of Finding. The template should clearly separate lesions, non-lesion changes, and special cases if these sections are used in the official edition.
• Lesion Morphology. Shape, orientation, margins, echotexture, and posterior acoustic features should be selected from a standardized list, not entered arbitrarily.
• Calcifications and Associated Features. Their presence is important not as a decorative line but as a factor influencing alignment with mammography and the final category.
• Additional Features. Vascularity and elastographic data should not replace the basic morphological assessment.
• Localization. A unified format for side, quadrant or clock face, distance from the nipple, and depth reduces errors during biopsy and dynamic observation.

BI-RADS 3: Why the 2% Number Remains Critical

BI-RADS 3 in ultrasound is permissible only when the finding truly corresponds to a probably benign model. The ≤2% threshold is not an “average risk by feel” but a limit separating observation from a suspicious category. Therefore, BI-RADS 3 should not be used as a compromise for an uncertain description or incomplete study.

If the doctor lacks data, BI-RADS 0 with further examination or comparison with previous studies is more appropriate. If the features exceed the probably benign range, the finding should move to BI-RADS 4 with a subcategory specified.

BI-RADS 4: Subcategories Must Be Retained in the Protocol

For ultrasound, detailing BI-RADS 4 is especially important. Category 4 covers a wide risk range — from >2% to <95%, so without 4A, 4B, or 4C, the conclusion is less informative for the clinician and patient. The subcategory helps align the ultrasound picture with mammography, MRI, clinical data, and choose the verification route.

Local templates should not leave a single “BI-RADS 4” button. It should expand to 4A, 4B, and 4C, otherwise one of the main functions of the system — stratification of suspicious findings — is lost.

How to Implement v2025 in the Ultrasound Department

1. Obtain the official text of the BI-RADS v2025 Manual and appoint a person responsible for comparing it with the old template.
2. Create a correspondence table: old term, new official term, permissible synonyms, prohibited free formulations.
3. Update structured protocols in the information system, including screening and diagnostic ultrasound.
4. Conduct a review of contentious cases of BI-RADS 3, 4A, 4B, 4C, and 5 at a local council.
5. Ensure that the final category automatically appears in the conclusion and does not contradict the description.

Short Conclusion for Practice

BI-RADS v2025 for breast ultrasound is an opportunity to update the protocol language, not just the category directory. The numerical risk framework remains key: ≤2% for BI-RADS 3, BI-RADS 4 gradations from >2% to <95%, ≥95% for BI-RADS 5. The full list of lexicon changes should be taken from the official ACR BI-RADS v2025 Manual, and local templates should be aligned with it without unofficial terminological additions.