Related Experiment Video
Updated: Jun 10, 2026
13:44
Detection of Architectural Distortion in Prior Mammograms via Analysis of Oriented Patterns
Published on: August 30, 2013
Estimating the mutual information between bilateral breast in thermograms using nonparametric windows.
M EtehadTavakol1, E Y K Ng, Caro Lucas
1Electrical and Computer Engineering Department, Isfahan University of Technology, Isfahan, Iran. m.etehadtavakol@ec.iut.ac.ir
Journal of Medical Systems
|August 13, 2010
Summary
Comparing contralateral breast thermal images helps detect breast cancer. Higher similarity between breast thermal images indicates lower abnormality risk, as shown by normalized mutual information values closer to one.
Area of Science:
- Medical Imaging
- Biomedical Engineering
- Radiology
Background:
- Breast cancer detection relies on comparing contralateral breast images.
- Asymmetric temperature distribution in thermal imaging can indicate abnormalities.
- Mutual information effectively measures nonlinear correlations for abnormality detection.
Purpose of the Study:
- To evaluate the effectiveness of normalized mutual information in detecting abnormalities in simulated breast thermal images.
- To assess the correlation between thermal image similarity and abnormality indicators.
Main Methods:
- Utilized sixty simulated breast thermal images.
- Employed nonparametric windows for enhanced image resolution.
- Calculated normalized mutual information to quantify image similarity and dependence.
Main Results:
- Demonstrated that higher similarity between left and right breast thermal images correlates with normalized mutual information values closer to one.
- Confirmed mutual information as a suitable indicator for detecting abnormalities.
Conclusions:
- Normalized mutual information is a reliable metric for assessing contralateral breast thermal image similarity.
- This method shows promise for non-invasive breast cancer detection through thermal imaging analysis.
