Chapter 8 - Deep similarity learning for disease prediction

Abstract

Deep learning is helping many medical experts as well as researchers to discover important insights from healthcare data and provide better medical facilities. Patients' risk assessment of developing a certain disease and providing personalized healthcare is an important research area. The current study presents a comprehensive review of deep learning architectures and how they can be used in learning representation and similarity between a pair of patients for disease prediction. We call this model a deep similarity learning model for disease prediction. As a demonstration of functionality, the deep learning architectures are trained on electronic health records to perform disease prediction. The experimental results obtained encourage us to use one of the suitable models in calculating similarity as future work.