An Optimization Algorithm for Computer-Aided Diagnosis of Breast Cancer Based on Support Vector Machine.

Area of Science:

• Oncology
• Computational Intelligence
• Biomedical Data Science

Background:

• Breast cancer incidence is rising in China, particularly among younger women, necessitating advanced risk assessment.
• Traditional diagnostic methods require enhancement for improved accuracy and efficiency.
• Data-driven statistical learning offers novel approaches to breast cancer diagnosis.

Purpose of the Study:

• To develop and evaluate an improved machine learning algorithm for breast cancer risk prediction and auxiliary diagnosis.
• To enhance the accuracy and efficiency of breast cancer diagnosis using computational intelligence.
• To explore the algorithm's applicability in classifying breast cancer across different stages.

Main Methods:

• Development of a hybrid optimization algorithm (GSP-SVM) combining genetic algorithm, particle swarm optimization, and simulated annealing with Support Vector Machine.
• Application of the GSP-SVM algorithm to historical breast cancer data for prediction and classification.
• Comparative analysis of GSP-SVM performance against other optimization algorithms using metrics like classification accuracy, MCC, and AUC.

Main Results:

• The GSP-SVM algorithm achieved high levels of classification accuracy, MCC, and AUC.
• The proposed method demonstrated superior performance compared to existing optimization algorithms.
• The algorithm showed potential for effective decision support in auxiliary breast cancer diagnosis.

Conclusions:

• The GSP-SVM algorithm offers a robust and effective tool for breast cancer auxiliary diagnosis, significantly improving diagnostic efficiency.
• The study highlights the potential of advanced machine learning techniques in addressing the growing challenge of breast cancer.
• Further exploration of the algorithm for multi-period and multi-class breast cancer detection is warranted.