Breast thermography as a screening technique. An evaluation of performance data

Area of Science:

• Medical Imaging
• Diagnostic Technology
• Biomedical Engineering

Background:

• Breast thermography performance evaluation is often anecdotal and lacks sufficient data.
• Conflicting conclusions exist regarding the efficacy of thermography in published literature.
• Standardized methods for performance assessment are needed.

Purpose of the Study:

• To introduce and apply relative operating characteristics (ROC) curves and the detectability index (d-prime) for evaluating breast thermography.
• To reconcile conflicting findings on thermography efficacy using quantitative performance metrics.
• To provide conclusions for optimizing current clinical thermographic techniques.

Main Methods:

• Utilized relative operating characteristics (ROC) curve analysis.
• Employed the detectability index (d-prime) for performance quantitation.
• Assessed and synthesized data from published breast thermography studies.

Main Results:

• Established that ROC curves and d-prime can effectively evaluate breast thermography.
• Demonstrated that quantitative analysis reconciles conflicting conclusions in existing literature.
• Identified performance metrics for assessing thermography efficacy.

Conclusions:

• Relative operating characteristics (ROC) curves and detectability index (d-prime) offer a robust framework for breast thermography performance evaluation.
• Quantitative analysis using d-prime can clarify the efficacy of thermography, resolving discrepancies in prior research.
• The findings provide a basis for optimizing existing clinical thermographic techniques and guiding future improvements.