CAD May Not be Necessary for Microcalcifications in the Digital era, CAD May Benefit Radiologists for Masses

Area of Science:

• Radiology
• Medical Imaging
• Oncology

Background:

• Early breast cancer detection is crucial for improved patient outcomes.
• Digital mammography is a standard screening tool.
• Computer-aided detection (CAD) systems aim to enhance radiologist performance.

Purpose of the Study:

• To evaluate the effectiveness of CAD in marking breast cancers on digital mammograms at diagnosis.
• To retrospectively assess CAD's ability to identify cancers on prior mammograms for earlier detection.
• To analyze factors influencing CAD marking, including breast density, mammographic views, and lesion characteristics.

Main Methods:

• Retrospective review of 3445 diagnosed breast cancers from 2004-2008, with 1293 imaged with full-field digital mammography (FFDM).
• Analysis focused on 43 cancers initially missed but visible on prior mammograms (false-negatives).
• Collected data included patient demographics, breast density, lesion characteristics, and CAD marking status on current and prior mammograms.

Main Results:

• CAD marked 79% of cancers at the time of diagnosis and 56% on prior mammograms.
• Effectiveness was consistent in dense breast tissue (79% at diagnosis, 56% on prior exams).
• CAD marked lesions on craniocaudal (CC) view more frequently (32%) than mediolateral oblique (MLO) view (21%) at diagnosis.

Conclusions:

• CAD demonstrates significant effectiveness in marking breast cancers at diagnosis and retrospectively on prior mammograms.
• CAD can identify invasive breast carcinomas, including those in dense breast tissue.
• The higher marking rate on the CC view suggests increased vigilance may be warranted when CAD highlights lesions on this specific view.