Related Experiment Video
Updated: Jul 10, 2026
13:44
Detection of Architectural Distortion in Prior Mammograms via Analysis of Oriented Patterns
Published on: August 30, 2013
Textural assessment in digital mammograms
F E Trujillo-Zamudio1, J Márquez, Y Villaseñor
1Inst. de Fisica, UNAM, Mexico City, Mexico. flaviotrujillo@gmail.com
Summary
This study evaluated digital mammogram features to detect cancer. Textural and morphological analysis showed limited diagnostic accuracy, correctly identifying cancer in 46% of cases.
Area of Science:
- Radiology and Medical Imaging
- Biomedical Engineering
- Oncology
Background:
- Digital mammography is crucial for breast cancer screening.
- Accurate interpretation of mammograms is essential for early cancer detection.
- BI-RADS 4 and 5 classifications indicate suspicious or malignant findings, necessitating further investigation.
Purpose of the Study:
- To assess the effectiveness of textural and morphological parameters in characterizing digital mammograms.
- To differentiate between cancerous and non-cancerous cases using image analysis techniques.
- To evaluate the diagnostic performance of proposed parameters in a clinical setting.
Main Methods:
- Analysis of digital mammograms from patients with BI-RADS 4 or 5 classifications.
- Utilizing textural roughness via the mean height-width ratio of extrema (MHWRE).
- Employing morphological feature analysis using circularity.
Main Results:
- The study achieved a correct diagnosis rate of 46% for the analyzed mammograms.
- Incorrect diagnoses were observed in 25% of the cases.
- A significant portion (29%) of diagnoses remained undetermined based on the tested parameters.
Conclusions:
- Textural and morphological parameters show potential but limited ability in definitively diagnosing breast cancer from mammograms.
- Further refinement of image analysis techniques is needed to improve diagnostic accuracy.
- The current parameters require enhancement for reliable clinical application in differentiating malignant findings.
