Related Experiment Video
Updated: Jul 31, 2026
13:44
Detection of Architectural Distortion in Prior Mammograms via Analysis of Oriented Patterns
Published on: August 30, 2013
A calibration approach to glandular tissue composition estimation in digital mammography
J Kaufhold1, J A Thomas, J W Eberhard
1General Electric Global Research Center, Niskayuna, New York 12309, USA. kaufhold@crd.ge.com
Medical Physics
|August 31, 2002
Summary
This study calibrates digital mammography systems to accurately measure breast tissue composition, improving breast cancer risk assessment by quantifying glandular tissue percentage pixel by pixel.
Area of Science:
- Medical Imaging
- Radiology
- Biophysics
Background:
- Breast composition, specifically glandular tissue, is linked to breast cancer risk.
- Previous methods for measuring glandular tissue were limited to relative measurements.
- Objective, quantitative analysis of breast tissue composition is needed for accurate risk assessment.
Purpose of the Study:
- To perform a careful calibration of a digital mammography system for quantitative breast tissue composition estimation.
- To develop a method for pixelwise estimation of percent glandular composition in patient breasts.
- To assess the impact of various error sources on the accuracy of tissue composition estimates.
Main Methods:
- Digital mammography system calibration using glandular-equivalent phantoms (0%, 50%, 100% glandular).
- Extraction of mean signal and noise levels to compute calibration curves.
- Application of the developed method to 23 digital mammograms for pixelwise composition estimation.
Main Results:
- Calibration curves were computed for quantitative tissue composition estimation.
- The method successfully estimated percent glandular composition on a pixelwise basis.
- Errors in compressed breast height estimation were the largest contributor to inaccuracy (+/-7% for 4 cm).
- Quantum noise was found to be a minor error source (<1% glandular).
Conclusions:
- A calibrated digital mammography system enables accurate, quantitative estimation of breast tissue composition.
- The developed method provides pixelwise glandular composition, aiding in breast cancer risk assessment.
- Understanding and quantifying error sources is crucial for reliable tissue composition analysis.
