Human-Centred Design of Low-Impact Sensing and Affective Technologies in Healthcare

In an era where technology increasingly permeates every aspect of human life, the way we interact with these technologies has become a crucial determinant of their success and adoption. While technological advancement often focuses on improving functionality and performance, the human experience of using these technologies has emerged as an equally critical consideration. This experience encompasses their intrusiveness, their demands on our attention, and their integration into our daily routines. The matter is particularly evident in healthcare and assistive technologies, where the effectiveness of a solution must be balanced against its impact on the user's quality of life. Traditional approaches to health monitoring and assistive technologies often prioritize accuracy and reliability over user experience. While such solutions provide valuable functionality, their intrusive nature can lead to reduced compliance, user fatigue, and ultimately, diminished effectiveness. The challenge therefore lies not just in creating functional technologies, but in developing solutions that work harmoniously with natural human behaviour and perception. Recent advances in sensing technology, artificial intelligence, and our understanding of human perception have opened new possibilities for more non-invasive approaches. However, these advances often develop in isolation, missing opportunities for synergistic integration. This thesis addresses these gaps through a multifaceted approach that combines theoretical understanding of human perception with practical applications in sensing and interface design. The research is driven by the hypothesis that understanding and incorporating human emotional and perceptual processes can lead to more effective and acceptable solutions in both healthcare monitoring and sensory assistance. In other words, it is driven by the hypothesis that human-centred design approaches to sensing and affective technologies can improve user acceptance and efficacy in healthcare applications. To this end, the work is structured around three interconnected research areas. The first investigates the role of emotion in crossmodal perception, establishing emotion as a mediator in audiovisual associations and developing quantitative frameworks for measuring emotional mediation. The second translates this theoretical understanding into emotion-aware technologies, advancing the state of the art through standardized validation methodologies for facial emotion recognition algorithms, emotion-congruent sensory substitution, and technology-enhanced therapeutic interventions. The third advances non-invasive sensing by reconsidering calibration strategies, measurement protocols, and analysis speed; specific contributions include optimised calibration for glucose monitoring, breath analysis for disease detection, and rapid spectrophotometric analysis for efficient screening. Together, these research areas form a comprehensive framework that builds on and extends the current state of the art in human-centred healthcare technologies. The scope of this research encompasses theoretical investigation, system development, and practical validation, and explicitly considers both the technical performance of these solutions and their impact on user experience, recognising that effective solutions must perform well on both dimensions. By examining the role of emotion in crossmodal perception, developing new approaches to affect-aware computing, and advancing non-invasive sensing solutions, this work contributes to multiple fields, including human-computer interaction, assistive technology, and healthcare monitoring. Ultimately, it contributes to the broader goal of developing technologies that enhance human capabilities while minimizing their impact on natural behaviour and experience.