Explainable Artificial Intelligence in Quantifying Breast Cancer Factors: Saudi Arabia Context

Area of Science:

• Medical Informatics
• Artificial Intelligence in Healthcare
• Oncology

Background:

• Breast cancer is the leading cancer among women in Saudi Arabia.
• Accurate prediction of benign versus malignant cases is crucial for effective treatment.
• Integrating advanced computational methods can improve diagnostic accuracy.

Purpose of the Study:

• To apply eXplainable Artificial Intelligence (XAI) techniques for breast cancer prediction in Saudi Arabian patients.
• To evaluate the performance of six machine learning models using clinical and pathological data.
• To enhance model interpretability using LIME and SHAP methods.

Main Methods:

• Six machine learning models were trained and evaluated on Saudi Arabian breast cancer data.
• Performance metrics included accuracy, precision, recall, F1 score, and AUC-ROC score.
• Local Interpretable Model-Agnostic Explanations (LIME) and SHAP were used for interpretability.

Main Results:

• The Random Forest model achieved the highest accuracy (0.72) and strong performance across other metrics.
• The Support Vector Machine model demonstrated the lowest predictive capability.
• XAI methods revealed varying feature importance across different models.

Conclusions:

• Machine learning, particularly Random Forest, shows promise for breast cancer prediction in Saudi Arabia.
• XAI techniques provide crucial insights into model decision-making processes.
• Findings support the potential integration of AI tools into clinical breast cancer diagnosis.