Deep Federated Learning for IoT to Improve Healthcare Operations

Document Type

Conference Proceeding

Publication Title

2023 International Conference on Intelligent Data Science Technologies and Applications, IDSTA 2023

Abstract

Recently, technological development has given rise to the Internet of Things (IoT), advanced machine learning techniques, wireless communications, and other technologies that have served as the basis for developing applications in various industries such as healthcare. Motivated by the need for a comprehensive and efficient framework that integrates these advanced technologies to take advantage of each of them, solve the problems of implementing each technology alone, and build a privacy-preserve, robust, and scalable solution. This paper presents a novel framework that applies Deep Federated Learning for healthcare IoT. This framework aims to improve healthcare operations, preserve patient privacy, and thus deliver better healthcare services. The proposed framework discusses the technical requirements required to implement such frameworks. Also, a thorough comparison with the existing literature is performed from a technical point of view. Furthermore, an experimental implementation of fetus health classification is presented, along with preliminary findings regarding the framework's ability to build high-performance predictive models considering the distributed nature of healthcare data and its limited availability. The results of improving health services by preventing potential deaths and disabilities are also presented.

First Page

85

Last Page

90

DOI

10.1109/IDSTA58916.2023.10317828

Publication Date

11-20-2023

Keywords

Deep Learning, Federated Learning, Healthcare, IoT

Comments

IR conditions: non-described

Link to Full Text

Share

COinS