A Web-based Telepsychology Platform Prototype Using Cloud Computing and Deep Learning Tools

Abstract

This paper describes a web-based telepsychology platform prototype, which enables patients and therapists to respectively follow and track psychological therapies remotely and asynchronously. The software architecture consists of two main modules: backend and frontend. The backend is designed by using several cloud tools, provided by Google, for processing and storing patients’ data in a secure manner and ensuring the highest privacy standards. In addition, therapies defined by therapists can also be uploaded to the platform for their posterior usage. The frontend is developed as a web application, which can be accessed according to the role of the user (patient or therapist), through a compatible browser. The presented prototype also allows some biomedical hardware devices to easily feed the platform with measurements that helps to estimate the stress degree and mood. Furthermore, the frontend includes a Deep Learning-based component capable of detecting facial landmarks, useful for estimating emotional features, which are stored preserving patients’ privacy. Finally, the results obtained after testing the prototype in a controlled environment are presented, together with a discussion about the advantages and drawbacks of such system.