Although everyday life seems to have returned to normal for most people after the COVID-19 pandemic, many are still battling long-term effects. Post-COVID syndrome (PCS), with its highly diverse and poorly understood symptoms, presents a challenge for clinicians and patients, as healthcare systems often face capacity limits to adequately address the individual needs of patients after diagnostics. In this work, we introduce a digital healthcare assistant based on a virtual agent (VA), serving as a low-barrier assistant that enables patients to continuously record their medical data regarding PCS using natural language interactions. The designs are focused on accessibility by supporting individual energy management mechanisms to enable pacing for those suffering from fatigue, as well as multilingual conversations to reduce language barriers.
Approximately three million adults in the Netherlands struggle with literacy, numeracy, and digital skills. This study investigates a GPT-4-based tool that simplifies Dutch healthcare documents using structured summaries and visual aids. In an exploratory study with native (NT1) and non-native (NT2) Dutch speakers, participants reported reduced cognitive load and anxiety, though comprehension gains were limited. The findings highlight the need for emotionally supportive interfaces, transparent AI outputs, and differentiated strategies based on literacy profiles. These insights support the design of inclusive, trustworthy AI tools that help low-literate adults access critical information.
Digital health innovations play an ever-growing role in health care and health self-management. Co-creation of digital health technology is a key method to increase its accessibility and use. However, many groups with complex care and support needs have been excluded from inclusive development and evaluation, such as people with mild intellectual disabilities (MID). This limits their ability to fully benefit from these technologies and exacerbates existing health inequities. This positioning paper draws upon the authors’ experience in designing and researching with people with MID and provides guidelines for inclusive design, co-creation, and evaluation of digital health with people with MID.
Elemental Allies beside You presents a mixed reality (MR) design concept that links walking, virtual companions, and public visualization to promote physical activity in socially meaningful ways. Instead of framing exercise as a repetitive task, the project reimagines walking as an embodied journey accompanied by collectible elemental sprites. Walking data such as step counts, routes, and temporal rhythms not only drive the appearance and evolution of these companions but also materialize in environmental visualizations that can be shared across community spaces. This paper positions walking as both embodied social interaction and as an opportunity for environment-based feedback, arguing that the combination of companionship, social encounters, and public visualization can provide intrinsic motivation and foster collective engagement in everyday physical activity.
Machine learning holds great promise for healthcare, but model transportability remains a key challenge to widespread adoption. Few studies use multi-center data, and even fewer address the heterogeneity such data introduces. This study evaluates the transportability of a model for detecting post-esophagectomy complications across two hospitals. Results highlight challenges due to differing treatment protocols, with models based on vital signs showing limited generalizability.
