An observer study for a computer-aided reading protocol (CARP) in the screening environment for digital mammography.

Area of Science:

• Radiology and Medical Imaging
• Health Informatics
• Workflow Optimization

Background:

• Digital mammography screening relies on efficient interpretation of images.
• Current protocols, like the quadrant view (QV), may not be optimal for workflow efficiency.
• Computer-aided detection (CAD) systems offer potential improvements.

Purpose of the Study:

• To evaluate the impact of a computer-aided reading protocol (CARP) on digital mammogram reading time and detection sensitivity.
• To compare CARP's efficiency and accuracy against the traditional quadrant view (QV) protocol.
• To determine if CARP can enhance workflow efficiency in mammography screening.

Main Methods:

• A receiver-operating characteristic (ROC) study involving 200 digital mammogram cases.
• Six radiologists interpreted cases using both QV and CARP workflows.
• Interpretation times and diagnostic performance were recorded and analyzed using ROC and kappa analysis.

Main Results:

• No statistically significant difference in reader performance (sensitivity) was observed between CARP and QV.
• A statistically significant reduction in average interpretation time for negative cases was found with CARP (58.8 seconds) compared to QV (64.7 seconds).
• Reader performance, measured by the area under the ROC curve, was comparable between the two protocols.

Conclusions:

• CARP, which prioritizes regions of interest based on CAD findings, can decrease interpretation time for negative mammograms.
• This reduction in reading time for negative cases, common in screening, can improve overall workflow efficiency.
• CARP offers a potential method to enhance mammography screening efficiency without compromising diagnostic sensitivity.