Breast Cancer Mammograms Classification Using Deep Neural Network and Entropy-Controlled Whale Optimization Algorithm

Area of Science:

• Medical Imaging
• Artificial Intelligence
• Computational Biology

Background:

• Breast cancer detection and classification are critical for early diagnosis and treatment, yet manual mammogram analysis by radiologists is time-consuming and challenging.
• Existing computer-aided diagnosis (CAD) systems require improvement for greater precision through advanced methods and technologies.
• There is a continuous need to enhance CAD systems to reduce false-positive rates and improve overall accuracy in breast cancer diagnosis.

Purpose of the Study:

• To investigate novel methods for breast cancer classification and prevention strategies.
• To enhance the accuracy of CAD systems by optimizing feature extraction and selection.
• To reduce the risk of breast cancer in women's lives through more precise diagnostic tools.

Main Methods:

• Deep feature extraction using fine-tuned MobilenetV2 and Nasnet Mobile architectures.
• Feature optimization and fusion using the proposed Modified Entropy Whale Optimization Algorithm (MEWOA).
• Classification of breast cancer images using machine learning classifiers on optimized deep features.

Main Results:

• The proposed MEWOA-based CAD system achieved high accuracy: 99.7% on the INbreast dataset, 99.8% on the MIAS dataset, and 93.8% on the CBIS-DDSM dataset.
• Significant improvement in CAD system accuracy was observed, primarily due to the reduction in false-positive rates.
• The proposed method demonstrated superior performance compared to existing approaches in breast cancer image classification.

Conclusions:

• The Modified Entropy Whale Optimization Algorithm (MEWOA) effectively enhances deep feature extraction and classification for breast cancer detection.
• The developed CAD system offers a more precise and efficient tool for radiologists, potentially leading to earlier diagnosis and better patient outcomes.
• The study highlights the potential of advanced AI algorithms in improving medical imaging analysis and reducing the global burden of breast cancer.