focalinterest.net

Tag: Digital Mental Health

  • Emotion-adaptive energy nudges improved engagement in a lab study

    What the study found

    The study found that emotionally adaptive energy-feedback nudges were associated with higher positive affect and sustained engagement than a non-adaptive baseline. The adaptive condition also increased exposure to conservation-relevant cues and produced modest gains in self-reported energy awareness.

    Why the authors say this matters

    The authors suggest that digital energy feedback is often limited when it uses static, uniform messages that ignore a user’s emotional context. They conclude that affect-aware adaptation may improve the long-term effectiveness of energy-conservation nudges.

    What the researchers tested

    The researchers proposed an affect-aware, model-free reinforcement learning framework for personalized energy-feedback nudging. The system extended the MAPE-K loop, which is a monitoring-and-adaptation framework, with an Affect–Behavior Decoupling Architecture that processed emotional signals from real-time facial emotion recognition and behavioral cues in parallel.

    What worked and what didn't

    In a simulated smart-home dashboard and a controlled laboratory study, the emotionally adaptive condition outperformed a non-adaptive baseline on positive affect and sustained engagement. It also increased exposure to conservation-relevant cues and led to modest gains in self-reported energy awareness. The abstract says that demonstrating direct impact on energy consumption still requires longitudinal field studies.

    What to keep in mind

    The study was done in a simulated smart-home dashboard and a controlled laboratory setting, so the results are not direct evidence of real-world energy savings. The abstract also notes that longitudinal field studies are needed to show direct impact on energy consumption.

    • Emotionally adaptive energy feedback was linked to higher positive affect and sustained engagement.
    • The system used real-time facial emotion recognition and behavioral cues to choose nudges.
    • The adaptive condition increased exposure to conservation-relevant cues.
    • Self-reported energy awareness rose modestly in the adaptive condition.
    • Direct effects on energy consumption were not demonstrated in this study.

  • Reasoning-based LLMs predicted 12-week antidepressant remission

    What the study found

    The study found that reasoning-based large language models could predict 12-week remission in patients with depressive disorder receiving antidepressant monotherapy. The best-performing model was Claude 3.7 Sonnet with 32,000 reasoning tokens and a referencing of deep research prompt.

    Why the authors say this matters

    The authors conclude that these models show promise as interpretable adjunctive tools in depressive disorder treatment planning. They also say prospective validation in real-world clinical settings remains essential.

    What the researchers tested

    The researchers analyzed data from 390 patients in the MAKE Biomarker discovery study who were taking first-step antidepressant monotherapy. They tested three large language models — ChatGPT o1, o3-mini, and Claude 3.7 Sonnet — using prompting strategies including zero-shot chain-of-thought, atom-of-thoughts, and a novel referencing of deep research prompt. Three psychiatrists independently rated the model outputs for clinical validity on 5-point Likert scales.

    What worked and what didn't

    Claude 3.7 Sonnet with 32,000 reasoning tokens and the referencing of deep research prompt achieved the highest performance, with balanced accuracy of 0.6697, sensitivity of 0.7183, and specificity of 0.6210. Medication-specific analysis showed negative predictive values of 0.75 or higher across major antidepressants, suggesting stronger performance for identifying likely nonresponders. Psychiatrists gave favorable mean ratings for correctness, consistency, specificity, helpfulness, and human likeness.

    What to keep in mind

    The study used retrospective data from a single biomarker discovery dataset after excluding patients with uncommon medications or missing biomarker data. The abstract does not describe longer-term follow-up beyond 12 weeks, and it states that prospective real-world validation is still needed.

    • The best model was Claude 3.7 Sonnet with 32,000 reasoning tokens and a referencing of deep research prompt.
    • Balanced accuracy for the top model was 0.6697, with sensitivity of 0.7183 and specificity of 0.6210.
    • Negative predictive values were 0.75 or higher across major antidepressants in medication-specific analysis.
    • Three psychiatrists rated the model outputs favorably on correctness, consistency, specificity, helpfulness, and human likeness.
    • The authors say prospective validation in real-world clinical settings remains essential.

  • Digital technologies support meaningful connections in care homes

    What the study found

    The review found that digital technologies in care homes have been used to support meaningful connections among residents, relatives, and staff. The technologies reported included robotics, virtual reality, mobile or tablet apps, digital devices, and online programs.

    Why the authors say this matters

    The authors say meaningful connections are important for social health and well-being, especially in care homes where barriers can increase social isolation. The study suggests digital technology can act as a catalyst for human connection rather than a replacement, and the authors say generative AI is a current gap that should be considered with key stakeholders.

    What the researchers tested

    The researchers carried out a scoping review of English-language studies on digital technologies used in long-term care settings to facilitate meaningful connections. They searched six databases, gray literature, and citation lists, and included studies involving care home residents, relatives, or staff that directly discussed a digital technology focused on building meaningful connections.

    What worked and what didn't

    Across 72 included studies, the review identified several factors linked to meaningful connections, including getting to know the person, increased autonomy and choice, enjoyment and fun, communication, and community building. The main indicators reported were engagement, well-being or satisfaction, emotional response, quality of life, purpose and meaning, social closeness, loneliness, depression and anxiety, and psychosocial capacity.

    What to keep in mind

    The authors note several limitations: the review was restricted to English-language studies, some studies on social connection may have been missed, passive technologies were excluded, and outcome measures were heterogeneous. The abstract also notes that the evidence base for generative AI in care homes is currently a gap.

    • The review included 72 studies on digital technologies used in care homes.
    • Reported technologies included robotics, virtual reality, mobile or tablet apps, digital devices, and online programs.
    • The authors say digital technology can support human connection rather than replace it.
    • Generative AI is described as a gap in the current evidence base.
    • The review found heterogeneous outcome measures across the included studies.