Contactless artificial intelligence-enabled radio frequency sensing for healthcare applications

Taylor, William Alexander (2023) Contactless artificial intelligence-enabled radio frequency sensing for healthcare applications. PhD thesis, University of Glasgow.

Abstract

Human motion detection is getting considerable attention in the field of Artificial Intelligence (AI) driven healthcare systems. Human motion detection can be used to provide remote healthcare solutions for vulnerable people by identifying particular movements such as falls, gait and breathing/heart disorders. This can allow people to live more independent lifestyles with the safety of being monitored if more direct care is needed. At present contact-based devices can provide real-time monitoring by deploying devices on a person’s body. However, placing devices on a person’s body all the time can be uncomfortable. Elderly people are also at risk of forgetting to wear devices. additionally, contact-based devices also require removal for recharging due to battery requirements. This thesis details the work that has been undertaken in the field of non-contact monitoring of human movements and vital signs. There is current research looking at using camera and radar technology to monitor vulnerable people within the home however, these techniques come with some disadvantages. Camera technology in the home has privacy concerns as people can feel uncomfortable being watched and radar technology introduces new technology into the home. This thesis explores the use of Radio Frequency (RF) signals to sense human movements and vital signs. RF signals are currently present in many homes as Wi-Fi networks already emit RF signals through the home. As people move around these signals, signal propagation is affected. Channel State Information (CSI) describes how a signal propagated from the transmitter to the receiver. This thesis takes the CSI and employs Machine Learning (ML) techniques to associate patterns observed in the CSI with specific movements. This principle is used to develop a real-time monitoring system that can detect what movements have occurred. The results of this thesis have shown to be able to differentiate between different activities using RF signals with over 90 % accuracy. This thesis serves as a proof of concept for contactless fall and vital sign detection systems that can assist elderly and vulnerable people to live independently without the need to wear monitoring devices.

• Item Type: Thesis (PhD)
• Qualification Level: Doctoral
• Subjects: T Technology > T Technology (General)
• Colleges/Schools: College of Science and Engineering > School of Engineering
• Supervisor's Name: Imran, Professor Muhammad Ali and Abbasi, Dr. Qammer
• Date of Award: 2023
• Depositing User: Theses Team
• Unique ID: glathesis:2023-83473
• Copyright: Copyright of this thesis is held by the author.
• Date Deposited: 10 Mar 2023 09:44
• Last Modified: 10 Mar 2023 09:55
• Thesis DOI: 10.5525/gla.thesis.83473
• URI: https://theses.gla.ac.uk/id/eprint/83473

Related URLs

• Article DOI
• Article DOI
• Article DOI
• Article DOI
• Article DOI
• Article DOI
• Article DOI
• Data DOI
• Data DOI