Personalized Care: A Clinical Decision Support System for Breast Cancer Screening Using Clustering and Classification

Abstract

In recent years, Computer-Aided Diagnosis (CADx) is popularly used in the healthcare system for the clinical guidelines. Breast cancer is the most leading cause of cancer death in women both in developed and developing countries. The lack of personalized care, early diagnosis and treatment of breast cancer is the major challenge for medical practitioners. To address this issue, personalized breast cancer screening is developed by considering the patient personal features as well as the screening features. This is to suggest the treatment and regular follow up. Therefore, a clinical decision support system is mandatory for the quality decision making and timely diagnosis for the women. The system entails the Partitioning around Medoids (PAM) clustering of personal features and active classifier on each cluster with the screening features. Breast Imaging Reporting and Data System (BI-RADS) scheme is used to place the findings from the screening into well defined categories. This system helps in reducing the cost of screening and providing the guidelines for the medical practitioners.

Keywords: Breast cancer and early detection, BI-RADS, Clinical decision support system, PAM clustering, Classifier