Now showing 1 - 2 of 2
  • Thumbnail Image
    Some of the metrics are blocked by your 
    Item type:Publication,
    STELLA: An Open-Source Platform for Auditable Voice-First Conversational Health Coaching Agents
    (ACM, 2026-07-21) ;
    Vinay, Rasita
    ;
    Deploying voice agents powered by LLMs creates tension between system auditability and the low latency required for natural conversation, while imposing heavy infrastructure burdens. We present STELLA, an open-source platform enabling the efficient deployment of voice-first agents. As a platform, STELLA provides ready-to-use infrastructure, including plan-based session management, modular expert pipelines, and a developer SDK. This eliminates the need for researchers to build complex audio and safety backends from scratch. To address the latency and auditability tension, STELLA’s default agent structurally decouples conversational presence from backend processing. Upon user input, the agent immediately delivers backchannel acknowledgments to maintain natural conversational flow, while safety arbitration and expert routing execute concurrently. A 32-participant field evaluation showed STELLA achieves 1.5-second median latency on commodity scientific hardware and above-median conversational quality in the 50th to 75th percentile ASAQ representative set, demonstrating feasibility for health coaching deployment and motivating future work on clinical applications such as cognitive stimulation therapy.
    Type:
    Journal:
  • Thumbnail Image
    Some of the metrics are blocked by your 
    Item type:Publication,
    Systematic scoping review and ethical analysis of digital health interventions for dementia and cognitive decline
    (Research Square (United States), 2026-08-19)
    Polat, Ceren
    ;
    ; ;
    Biller-Andorno, Nikola
    ;
    Vinay, Rasita
    Digital health interventions (DHIs) support people with dementia, mild cognitive impairment (MCI), and subjective cognitive decline (SCD), but evidence and ethical reporting are fragmented. We searched five databases on 9 October 2025 and included 113 studies evaluating DHIs with standardised, objective, or psychometrically validated outcomes. We mapped study designs, populations, intervention focuses, delivery modes, author-reported safeguards, and potential ethical concerns identified through a four-principles-informed framework. VR/AR/MR (49.6%) and robotics (26.5%) were most common; cognitive (34.5%) and physical–cognitive activities (23.0%) predominated. MCI (56.6%) and dementia (54.9%) were frequently represented, whereas SCD appeared in 5.3% of studies. Reported safeguards emphasised physical and psychological safety and data anonymisation; potential concerns included digital literacy, unequal access, consent over time, and emotional attachment. The evidence base was technologically diverse but uneven across populations and ethical domains, identifying priorities for transparent reporting and person-centred DHI design.
    Type:
    Journal: