A Survey of Convolutional Neural Network in Breast Cancer

Area of Science:

• Oncology
• Medical Imaging
• Artificial Intelligence

Background:

• Breast cancer is a leading cause of death globally and the most common cancer in women.
• Early diagnosis significantly improves treatment options, effectiveness, and survival rates for breast cancer patients.
• Computer-aided diagnosis (CAD) systems are increasingly utilized for breast cancer detection.

Purpose of the Study:

• To conduct a comprehensive review of breast cancer diagnosis using convolutional neural networks (CNNs).
• To explore the application of CNNs in various breast cancer diagnostic tasks, including classification, detection, and segmentation.
• To identify current limitations and future research directions for CNN-based breast cancer diagnosis.

Main Methods:

• Review of recent scientific literature on CNNs for breast cancer diagnosis.
• Introduction to various medical imaging modalities used in breast cancer diagnosis.
• Detailed explanation of CNN architecture and its relevance to medical image analysis.
• Discussion of public breast cancer datasets and their characteristics.
• Categorization of CNN applications into classification, detection, and segmentation tasks.

Main Results:

• CNN-based approaches have demonstrated significant success in breast cancer diagnosis.
• The review covers diverse CNN applications across classification, detection, and segmentation tasks.
• Identified limitations include data scarcity, computational demands, and overfitting issues.

Conclusions:

• CNNs show great promise for advancing breast cancer diagnosis.
• Addressing limitations in dataset quality and size is crucial for further development.
• Future research should focus on improving CNN model efficiency and generalizability for robust clinical application.