Advancing electronic healthcare in ambient assistive living utilizing intelligent mobile systems

Koulouris, Dionysios; Κουλούρης, Διονύσιος

Doctoral Thesis

Author

Koulouris, Dionysios

Κουλούρης, Διονύσιος

Date

2025

Abstract

The continuously arising demand for advanced healthcare services puts significant pressure on current solutions, especially in the field of ambient assisted living. Such solutions include traditional systems which often incorporate static, expensive and not easy-to-use hardware. Remote patient monitoring, rehabilitation, physical activity motivation, navigation and accessibility are key scientific problems that the existing healthcare landscape tries to address using the mentioned conventional methods. The latest technological advancements provide increased computational capabilities in even smaller devices, more lasting batteries and faster interconnectivity. In combination with the continuously evolving sensors (cameras, gyroscopes, accelerometers), mobile devices offer a set of data which can be efficiently utilized to cover various scenarios. As a result of these innovations, demanding Machine Learning (ML) algorithms are now capable of being executed in the palm of the user hands. Hence, tasks such as Computer Vision (CV), a subfield of Artificial Intelligence (AI) can be efficiently used in mobile devices with applications like Augmented Reality (AR) offering overlay of computer-generated content in the real world. In addition, Human-Machine Interaction plays a pivotal role in user engagement as systems with such advanced principles are proven to offer satisfactory experiences for the users.This dissertation aims to address the aforementioned problem of advancing ambient assisted living by utilizing intelligent mobile systems such as mobile phones and smart watches. This thesis divides this goal into research questions including the feasibility of commodity devices to achieve the thesis goal, the importance of Human-Computer Interaction principles in healthcare, the integration of AR, IoT and Sensors technologies for remote patient monitoring, the increase in accessibility via intelligent handheld solutions, the location-aware possibilities to increase quality of life and medical education. Therefore, novel solutions are introduced along with their results which highlight a direction towards a more pervasive healthcare landscape. These solutions include mobile patient monitoring systems which utilize gamification techniques, AR and IoT to motivate users to increase their physical activity indoors, to visit pocket parks and to utilize onboard motion sensors to monitor their gait asymmetries. Moreover, intelligent AR algorithms are used to achieve obstacle avoidance for the visually impaired and increase accessibility in exhibition areas. Handheld location-awareness is examined in the field of indoor localization, providing solutions to achieve indoor navigation without the need of additional hardware, incorporated by ML, and to facilitate better crowd management in museum scenarios. Finally, the combination of AR is combined with CV in commodity mobile devices to create a medicine box identification and compound visualization system, as a proof-of-concept prototype of a medical education assistant.Overall, this thesis highlights that commodity devices, which have the benefits of cost and hardware elimination, can indeed facilitate an advanced ambient assisted living. Intelligent systems running on these devices are getting advantage of the onboard hardware and the edge to introduce novel proof-of-concept solutions and propose a direction towards accessible, personalized and cost-effective healthcare.

Department

Σχολή Τεχνολογιών Πληροφορικής και Επικοινωνιών. Τμήμα Ψηφιακών Συστημάτων

Number of pages

226

Language

English

URI

https://dione.lib.unipi.gr/xmlui/handle/unipi/18620

Collections

Τμήμα Ψηφιακών Συστημάτων

Show full item record

Except where otherwise noted, this item's license is described as
Αναφορά Δημιουργού-Όχι Παράγωγα Έργα 3.0 Ελλάδα