Breast density in dedicated breast computed tomography: Proposal of a classification system and interreader...

Area of Science:

• Radiology
• Medical Imaging
• Breast Imaging

Background:

• Mammography's ACR BI-RADS scale classifies breast density, impacting lesion detection.
• Dedicated breast computed tomography (BCT) requires a similar density classification system.
• Current BCT analysis lacks a standardized density assessment correlating with lesion visibility.

Purpose of the Study:

• To develop a novel breast density classification system for BCT, termed BCT density (BCTD).
• To base the BCTD system on lesion detectability, analogous to mammography's BI-RADS scale.
• To evaluate the interrater reliability of the proposed BCTD classification system.

Main Methods:

• A retrospective analysis of 114 contrast-enhanced BCT examinations was performed.
• An atlas-based BCTD classification system with 4 categories was defined, focusing on lesion detectability.
• Interrater reliability was assessed using 40 BCT datasets evaluated by 3 experienced radiologists.

Main Results:

• The BCTD system identified the lack of fatty septae as crucial for lesion detection in glandular tissue.
• A 10 mm lesion diameter cut-off was established for recommending additional ultrasound.
• The BCTD classification showed high interrater reliability, with an intraclass correlation coefficient (ICC) of 0.85–0.87.

Conclusions:

• A dedicated atlas-based BCTD classification system calibrated to lesion detectability has been proposed.
• The new BCTD system demonstrates high interrater reliability.
• This BCTD system can assist in decisions regarding the need for additional ultrasound examinations.