Utilizing offline reinforcement learning (RL) with real-world clinical data is getting increasing attention in AI for healthcare. However, implementation poses significant challenges. Defining direct rewards is difficult, and inverse RL struggles to infer accurate reward functions from expert behavior in complex environments. Offline RL also encounters distributional discrepancies between learned policies and observed human behavior, a critical issue in healthcare applications. To address challenges in applying offline RL to physical activity promotion for older adults at high risk of falls, based on wearable sensor activity monitoring, we introduce Kolmogorov-Arnold Networks and Diffusion Policies for Offline Inverse Reinforcement Learning (KANDI). Specifically, by leveraging the flexible function approximation in Kolmogorov-Arnold Networks, we estimate the reward function by learning free-living environment behavior from low-fall-risk older adults (experts). Additionally, diffusion-based policies within an Actor-Critic framework provide a generative approach for action refinement, enabling controlled exploration and mitigating distributional shift issues in offline RL. We evaluate KANDI using wearable activity monitoring data in a two-arm clinical trial from our Physio-feedback Exercise Program (PEER) study, emphasizing its practical application in a fall-risk intervention program to promote physical activity among older adults. Our analysis identifies the optimal timing for anti-sedentariness interventions tailored to varying levels of fall risk, thereby maximizing daily physical activity. Additionally, we evaluate KANDI on the D4RL benchmark, outperforming the state-of-the-art methods in each domain. These results underscore KANDI's potential to address key challenges in offline RL for healthcare applications, offering an effective solution for the optimal timing and policy for activity promotion intervention strategies.

With the rise of the aging population, the demand of care increases yet the turnover rate of caregivers for geriatric patients has increased over the past few decades. While many factors contribute to the caregivers’ turnover rate, inadequate training and lack of communication skills and relationship-building skills are vital issues. Technology can help in providing training for soft skills. Embodied Virtual Human (VH) assistants and Augmented Reality (AR) can be a key method to train caregivers and to improve their experience in interacting with older adults, and their perception of the interaction. We designed and developed an immersive simulation in AR where caregivers interact with an embodied VH in two different conditions, i.e., first with an unaware VH and then with a VH having simulated awareness achieved through human-in-the-loop. We conducted a study with caregivers of older adults to evaluate the usability and technology acceptance of our system. The findings suggest that the majority of the participants found the system acceptable, and rated the system as user-friendly and easy to use.

The existing educational programs for caregivers often fall short in providing adequate training for essential interpersonal skills, particularly in communication and relationship development. To address this gap, researchers have highlighted the potential of Intelligent Virtual Agents (IVAs) as a tool for soft skills training across various sectors, including healthcare. In this article, we introduce an aware IVA (A-IVA) designed to represent a geriatric patient, that demonstrates limited and simulated awareness of the environment and previous conversations achieved by Wizard-of-Oz. We developed a novel Augmented Reality (AR) simulation that immerses users in a realistic caregiving scenario. We conducted a pilot study to evaluate how the A-IVA’s awareness capabilities influence geriatric caregivers’ social presence and empathy levels as compared to an unaware IVA (U-IVA).