A Bottom-Up Review of Image Analysis Methods for Suspicious Region Detection in Mammograms

Area of Science:

• Medical Imaging
• Radiology
• Artificial Intelligence in Healthcare

Background:

• Breast cancer is a leading cause of death in women globally, making early detection crucial for survival.
• Mammography is a key imaging modality, but its accuracy is challenged by manual interpretation and variations in breast density and mass features.
• Radiologists face a high workload analyzing numerous mammograms, increasing the risk of oversight errors.

Purpose of the Study:

• To provide a comprehensive survey of methodologies for detecting suspicious regions in mammograms.
• To assist radiologists and practitioners with tools for improved breast cancer detection.
• To explore techniques ranging from traditional image features to advanced AI-based approaches.

Main Methods:

• A bottom-up review of scientific methodologies for suspicious region detection in mammograms.
• Analysis of techniques based on low-level image features.
• Inclusion of recent advancements in AI-based approaches for computer-aided mass detection.

Main Results:

• The paper details various techniques, highlighting their theoretical and practical pros and cons.
• It covers a spectrum of methods from basic image processing to sophisticated AI models.
• Datasets and strategies relevant to mammogram analysis are discussed.

Conclusions:

• Computer-aided detection systems can serve as valuable second opinion tools for radiologists.
• A thorough understanding of diverse techniques is essential for advancing breast cancer detection accuracy.
• The survey aims to equip readers with comprehensive knowledge on mammogram analysis strategies and datasets.