Data Security Challenges in Deep Neural Network for Healthcare IoT Systems

Ho, Edmond S.L. (2022) Data Security Challenges in Deep Neural Network for Healthcare IoT Systems. In: Security and Privacy Preserving for IoT and 5G Networks: Techniques, Challenges, and New Directions. Studies in Big Data, 95 . Springer, Cham, Switzerland, pp. 19-37. ISBN 9783030854270, 9783030854300, 9783030854287

Preview

Text _IoT5GCyber_final_Data_Security_Challenges_in_Deep_Neural_Network_for_Healthcare_IoT_Systems.pdf - Accepted Version Download (8MB) | Preview

Abstract

With the advancement of IoT technology, more and more healthcare applications were developed in recent years. In addition to the traditional sensor-based systems, image-based healthcare IoT systems become more popular since no specialized sensors are required. Combining with Deep Neural Network (DNN) based automated diagnosis and decision-making systems, it is possible to provide users with 24/7 health monitoring in real life. However, the high computational cost for training DNNs can be a hurdle for developing such kind of powerful systems. While cloud computing can be a feasible solution, uploading training data for the DNN models to the cloud may lead to data security issues. In this chapter, we will review some image-based healthcare IoT systems and discuss some potential risks on data security when training the DNN models on the cloud.

Actions (login required)

View Item

Downloads