[Research progress of breast pathology image diagnosis based on deep learning]

Area of Science:

• Oncology
• Medical Imaging
• Artificial Intelligence

Background:

• Breast cancer diagnosis relies on histopathological images.
• Deep learning shows promise in medical image analysis.
• Traditional methods face limitations in breast cancer pathology classification.

Purpose of the Study:

• To review deep learning applications in breast pathology image analysis.
• To highlight advances in multi-scale feature extraction, cellular analysis, and classification.
• To discuss multimodal data fusion and future prospects in deep learning for breast cancer diagnosis.

Main Methods:

• Review of current deep learning techniques applied to breast pathology images.
• Focus on multi-scale feature extraction, cellular feature analysis, and classification.
• Summary of multimodal data fusion strategies for enhanced diagnostic accuracy.

Main Results:

• Deep learning outperforms traditional methods in breast cancer pathology classification.
• Multi-scale and cellular feature analysis are key areas of deep learning advancement.
• Multimodal data fusion offers advantages for comprehensive breast pathology image analysis.

Conclusions:

• Deep learning is a powerful tool for breast cancer pathology image diagnosis.
• Further research into challenges and future prospects will advance clinical applications.
• This review provides guidance for integrating deep learning into breast cancer diagnostic workflows.