Test Sensitivity in the Computer-Aided Detection of Breast Cancer from Clinical Mammographic Screening: a Meta-analysis

TL;DR

This meta-analysis evaluates how test sensitivity is measured in mammographic screening studies with CAD, revealing that improper comparisons may underestimate CAD benefits, but overall, CAD systems aid breast cancer detection.

Contribution

It highlights methodological issues in sensitivity measurement in CAD studies and shows that proper analysis confirms CAD's effectiveness in breast cancer screening.

Findings

Inappropriate sensitivity comparisons can underestimate CAD benefits.

Proper analysis supports CAD systems' positive role in screening.

Two large studies are significantly impacted by measurement issues.

Abstract

Objectives: To assess evaluative methodologies for comparative measurements of test sensitivity in clinical mammographic screening trials of computer-aided detection (CAD) technologies. Materials and Methods: This meta-analysis was performed by analytically reviewing the relevant literature on the clinical application of computer-aided detection (CAD) technologies as part of a breast cancer screening program based on x-ray mammography. Each clinical study's method for measuring the CAD system's improvement in test sensitivity is examined in this meta-analysis. The impact of the chosen sensitivity measurement on the study's conclusions are analyzed. Results: This meta-analysis demonstrates that some studies have inappropriately compared sensitivity measurements between control groups and CAD enabled groups. The inappropriate comparison of control groups and CAD enabled groups can lead to an underestimation of the benefits of the clinical application of computer-aided detection technologies. Conclusions: The potential for the sensitivity measurement issues raised in this meta-analysis to alter the conclusions of multiple existing large clinical studies is discussed. Two large scale studies are substantially affected by the analysis provided in this study and this meta-analysis demonstrates that computer-aided detection systems are successfully assisting in the breast cancer screening process.