Trustworthy Deep Feature Extraction and Ensemble-Based Machine Learning Approach for Breast Cancer Detections

Area of Science:

• Oncology
• Medical Imaging
• Artificial Intelligence in Medicine

Background:

• Breast cancer (BC) is a leading cause of cancer-related deaths globally.
• Current BC detection methods face challenges in combining high accuracy with interpretability, especially with complex imaging data.
• Diagnostic accuracy can be subjective and influenced by the clinician's expertise.

Purpose of the Study:

• To develop a reliable breast cancer detection strategy by integrating deep learning (DL) and ensemble machine learning (ML) techniques.
• To enhance the accuracy and interpretability of breast cancer diagnosis from medical images.
• To improve clinical decision-making and patient outcomes in breast cancer care.

Main Methods:

• Utilized a pre-trained deep learning model for effective feature extraction from breast cancer images.
• Applied eight different machine learning models for breast cancer identification.
• Evaluated model performance using precision, recall, F1-score, confusion matrices, and ROC curves.

Main Results:

• Achieved a high accuracy rate of 97.50%, with precision at 97.15%, recall at 97.00%, and F1-score at 96.98%.
• Demonstrated superior performance compared to existing state-of-the-art breast cancer detection models.
• Identified the support vector classifier, when combined with the pre-trained VGG-16 architecture, as the most effective ML model.

Conclusions:

• The proposed hybrid DL-ML strategy offers a significant advancement in breast cancer detection.
• The approach provides a reliable and accurate method for identifying breast cancer, aiding clinical decisions.
• This research contributes to improved patient care and better breast cancer outcomes through enhanced diagnostic tools.