Machine Learning Methods for Computer-Aided Breast Cancer Diagnosis Using Histopathology: A Narrative Review

Area of Science:

• Pathology
• Medical Imaging
• Artificial Intelligence

Background:

• Histopathology is a cornerstone of breast cancer diagnosis.
• Machine learning (ML) shows promise in medical image analysis.
• Conventional microscopy remains a standard for histopathological examination.

Purpose of the Study:

• To investigate the impact and application of ML in breast cancer diagnosis using conventional histopathology images.
• To review different ML approaches for image preprocessing, feature extraction, and classification in this context.
• To identify research gaps and propose future directions for ML in real-world pathology settings.

Main Methods:

• Review of existing literature on ML applications for breast cancer diagnosis via histopathology.
• Analysis of various techniques for image preprocessing, feature extraction, and classification.
• Discussion of deep learning's prevalence and potential for conventional microscopy.

Main Results:

• Most ML research in breast cancer diagnosis focuses on deep learning.
• ML methods offer significant potential to improve conventional microscopy-based diagnosis.
• Several research gaps exist for implementing ML in practical pathology environments.

Conclusions:

• ML holds considerable promise for advancing breast cancer diagnosis using conventional histopathology.
• Further research is needed to bridge the gap between ML development and real-world pathology implementation.
• Future guidelines should focus on integrating ML effectively into routine diagnostic workflows.