Machine Learning-Based Approaches for Breast Density Estimation from Mammograms: A Comprehensive Review.

Area of Science:

• Radiology and Medical Imaging
• Oncology
• Artificial Intelligence in Healthcare

Background:

• Breast cancer is the most common cancer in women, with breast density being a significant risk factor.
• High breast density can obscure tumors on mammograms, reducing detection sensitivity.
• Accurate breast density assessment is vital for risk stratification and early diagnosis.

Purpose of the Study:

• To conduct a comprehensive review of mammographic density estimation techniques.
• To focus on machine learning-based approaches for breast density assessment.
• To identify current limitations and suggest future research directions.

Main Methods:

• Review of visual, software-, machine learning-, and segmentation-based methods for mammographic density estimation.
• Categorization of machine learning methods into traditional and deep learning approaches.
• Analysis of commonly used models like Support Vector Machines (SVMs) and Convolutional Neural Networks (CNNs).

Main Results:

• Machine learning models, especially SVMs and CNNs, achieve high classification accuracies (76.70%–98.75%).
• Current methods are limited by subjectivity and cost-inefficiency.
• Deep learning models demonstrate significant potential in mammographic density estimation.

Conclusions:

• Machine learning offers powerful tools for objective and efficient mammographic density estimation.
• Future research should address subjectivity and cost barriers, exploring unsupervised segmentation and advanced models like transformers.
• Improved breast density estimation methods can enhance early breast cancer detection and diagnosis.