Using a human visual system model to optimize soft-copy mammography display: influence of MTF compensation

Area of Science:

• Medical Imaging
• Radiology
• Human Visual System Modeling

Background:

• Digital mammography relies on display technology like cathode ray tube (CRT) monitors.
• CRT monitor performance can be limited by modulation transfer function (MTF) deficiencies.
• Optimizing image display is crucial for accurate microcalcification detection.

Purpose of the Study:

• To develop an efficient method for optimizing CRT monitor performance in digital mammography.
• To correlate human observer performance with a mathematical model of the human visual system.
• To validate a human visual system model using observer performance data.

Main Methods:

• Six radiologists evaluated 250 mammographic images with varying microcalcification contrast levels on a CRT monitor.
• Images were viewed both unprocessed and processed to compensate for CRT MTF deficiencies.
• The JNDmetrix Visual Discrimination Model was used, and receiver operating characteristic (ROC) curves compared human and model observer performance.

Main Results:

• Both human and model observers showed improved performance with MTF-compensated images, particularly for mid-contrast microcalcifications.
• A very high correlation was observed between human and model observer performance.
• The study validated the predictive accuracy of the human visual system model.

Conclusions:

• Image processing to compensate for CRT MTF limitations enhances radiologist detection of mammographic microcalcifications.
• A human visual system-based model can accurately predict human observer performance.
• This approach aids in optimizing display systems for digital mammography.