Android Based Naive Bayes Probabilistic Detection Model for Breast Cancer and Mobile Cloud Computing: Design and Implementation

Abstract

Mobile phone technology initiatives are revolutionizing healthcare delivery in Africa and other developing countries. M-health services have transformed maternal health, management of communicable diseases such as Ebola and prevention of chronic diseases. Technological innovations in m-health have improved healthcare efficiency and effectiveness as well as extending health services to remote locations in rural African communities. This paper describes a ubiquitous m- health system that is based on the user centric paradigm of Mobile Cloud Computing (MCC) and android medical-data mining techniques. The development of ultra-fast 4G mobile networks and sophisticated smartphones and tablets has brought the cloud computing paradigm to the mobile domain.The system’s client side is based on an android platform for breast bio-data collection; a data mining technique based on Naïve Bayes probabilistic classifier (NBC) algorithm for predicting malignancy in breast tissue and the server-side MCC data storage. Experimental results indicate that the android Naïve Bayes classifier achieves 96.4% accuracy on Wisconsin Breast Cancer (WBC) data from UCI machine learning database.