This study presents a novel transfer learning approach and data augmentation technique for mental stability classification using human voice signals and addresses the challenges associated with limited data availability. Convolutional neural networks (CNNs) have been employed to analyse spectrogram images generated from voice recordings. Three CNN architectures, VGG16, InceptionV3, and DenseNet121, were evaluated across three experimental phases: training on non-augmented data, augmented data, and transfer learning. This proposed transfer learning approach involves pre-training models on the augmented dataset and fine-tuning them on the non-augmented dataset while ensuring strict data separation to prevent data leakage. The results demonstrate significant improvements in classification performance compared to the baseline approach. Among three CNN architectures, DenseNet121 achieved the highest accuracy of 94% and an AUC score of 99% using the proposed transfer learning approach. This finding highlights the effectiveness of combining data augmentation and transfer learning to enhance CNN-based classification of mental stability using voice spectrograms, offering a promising non-invasive tool for mental health diagnostics.
Antonis Psistakis, Burak Ocalan, Fabien Chaix
et al.
Ensuring resilience in distributed systems has become an acute concern. In today's environment, it is crucial to develop light-weight mechanisms that recover a distributed system from faults quickly and with only a small impact on the live-system throughput. To address this need, this paper proposes a new low-overhead, general recovery scheme for modern non-transactional leaderless distributed systems. We call our scheme HEAL. On a node failure, HEAL performs an optimized online incremental recovery. This paper presents HEAL's algorithms for settings with Linearizable consistency and different memory persistency models. We implement HEAL on a 6-node Intel cluster. Our experiments running TAOBench workloads show that HEAL is very effective. HEAL recovers the cluster in 120 milliseconds on average, while reducing the throughput of the running workload by an average of 8.7%. In contrast, a conventional recovery scheme for leaderless systems needs 360 seconds to recover, reducing the throughput of the system by 16.2%. Finally, compared to an incremental recovery scheme for a state-of-the-art leader-based system, HEAL reduces the average recovery latency by 20.7x and the throughput degradation by 62.4%.
Abstract This umbrella review explores the effects of Extended Reality (XR) and Game-Based Interventions (GBI) on anxiety, depression, and stress, covering augmented (AR), virtual (VR), and mixed reality (MR), along with serious games, gamification, game-based learning and training, exergames, and commercial video games. Following PRISMA and AMSTAR 2 guidelines, 201 articles were screened, with 16 reports selected (nine meta-analyses, six systematic reviews, and one scoping review). Findings highlight XR-GBI’s potential as a promising, flexible, and replicable intervention, demonstrating significant preliminary mental health improvement across children, adolescents, adults, and older people. Regarding anxiety, VR aids preoperative and procedural anxiety, promotes distraction and relaxation, and supports VR exposure therapy (VRET), matching Cognitive Behavioral Therapy (CBT) effectiveness with higher engagement. For depression, VRET reduces symptoms, while VR exergames combining physical activity and engagement show strong antidepressant effects. Stress management remains less explored, though AR video games enhance cognitive and social well-being, and VRET alleviates stress symptoms. Despite the XR-GBI promise, research is still emerging, with publications only beginning to expand recently, few randomized controlled trials, and methodological limitations. From our findings, we highlight practical and theoretical implications by showing how XR-GBI rely on core technical features and proposing a five-pathway theoretical model (cognitive, emotional, bodily, social, and motivational) that systematizes their potential for mental health, guiding future design, evaluation, and research. Further research should also expand on AR, MR, gamification, game-based learning and training, biofeedback, neurophysiological assessment, and social dynamics, while integrating artificial intelligence, digital mental health literacy, and psychoeducation to enhance XR-GBI’s impact.
Andria B. Eisman, Christine Koffkey, Judy Fridline
et al.
Background School-based universal prevention programs, like the Michigan Model for Health™ (MMH), hold promise for enhancing youth behavioral health but often face implementation challenges due to insufficiently addressing priority student issues. Previous research identified trauma-sensitive content as a student need in the MMH. Enhanced Replicating Effective Programs (REP), a multicomponent implementation strategy, is well suited to support program providers in addressing priority health issues among youth. Method This pilot cluster-randomized controlled trial compared Enhanced REP (tailored curriculum, training, and implementation facilitation with trauma-sensitive content) to standard REP (standard curriculum, initial training, as-needed technical assistance) across eight high schools serving low-income students. Through semistructured interviews at three time points, we assessed teacher perceptions of feasibility, acceptability, and appropriateness related to REP core and enhanced components. Results Teachers generally found Enhanced REP to deliver MMH satisfactory and suitable. However, the school environment, notably administrative support, influenced feasibility compared to standard REP. Enhanced REP teachers reported benefits in meeting student needs that were not seen in the standard REP group. The standard REP data helped to understand the comparative value of the enhanced strategy during a time of notable upheaval and mental health challenges due to the COVID-19 pandemic. Conclusions While some schools may succeed with less intensive strategies (REP), many may require more intensive approaches for effective implementation. Enhanced REP shows promise in tailoring curriculum delivery and providing additional support to meet student needs, but its success may hinge on organizational support, especially from leadership. Future research should investigate the addition of organizational-level strategies, such as leadership training, to optimize implementation and explore the comparative effectiveness of Enhanced versus standard REP.
Jo-anne Hughson, Mary Belfrage, Harold Douglas
et al.
Objectives: Dementia is a global health issue. Although best-practice guidelines for detection and management of dementia exist for primary care, there is a pressing need for culturally appropriate resources to support care for Aboriginal and Torres Strait Islander people living with cognitive impairment and dementia. Methods: A best-practice guide (BPG) for cognitive and dementia care for Aboriginal and Torres Strait Islander people attending primary care was developed incorporating evidence-based clinical care recommendations and cultural dimensions. Adopting research approaches characterised by inclusivity and collaboration, guide development included: (i) a planning phase with stakeholders; (ii) literature review; (iii) a draft development phase; (iv) a modified Delphi (e-Delphi) process; (v) in-depth cultural review by the project's Indigenous Reference Group and health service staff; (vi) and final clinical review. Results: Stakeholders wanted a BPG that was: comprehensive, easy to follow and practical; incorporated cultural considerations and; took account of the ongoing effects of colonisation. A two-part e-Delphi process, completed by 39 and 31 purposively selected participants respectively, reached consensus on: guide aims; cultural principles for inclusion; detection processes; future planning; referral to specialist cognition and palliative care services. Cultural review resulted in further integration of cultural principles and recommended development of additional resources. Further minor modifications were made during the final clinical review for peak body endorsement. Discussion: A rigorous development and review process has resulted in a culturally adapted resource health professionals can use to guide care with Aboriginal and Torres Strait Islander patients at risk of or experiencing cognitive impairment or dementia.
Theory of Mind (ToM), the ability to understand people's minds based on their behavior, is key to developing socially intelligent agents. Current approaches to ToM reasoning either rely on prompting Large Language Models (LLMs), which are prone to systematic errors, or use handcrafted, rigid agent models for model-based inference, which are more robust but fail to generalize across domains. In this work, we introduce AutoToM, an automated agent modeling method for scalable, robust, and interpretable mental inference. Given a ToM problem, AutoToM first proposes an initial agent model and then performs automated Bayesian inverse planning based on this model, leveraging an LLM backend. Guided by inference uncertainty, it iteratively refines the model by introducing additional mental variables and/or incorporating more timesteps in the context. Across five diverse benchmarks, AutoToM outperforms existing ToM methods and even large reasoning models. Additionally, we show that AutoToM can produce human-like confidence estimates and enable online mental inference for embodied decision-making.
The early detection of mental health disorders from social media text is critical for enabling timely support, risk assessment, and referral to appropriate resources. This work introduces multiMentalRoBERTa, a fine-tuned RoBERTa model designed for multiclass classification of common mental health conditions, including stress, anxiety, depression, post-traumatic stress disorder (PTSD), suicidal ideation, and neutral discourse. Drawing on multiple curated datasets, data exploration is conducted to analyze class overlaps, revealing strong correlations between depression and suicidal ideation as well as anxiety and PTSD, while stress emerges as a broad, overlapping category. Comparative experiments with traditional machine learning methods, domain-specific transformers, and prompting-based large language models demonstrate that multiMentalRoBERTa achieves superior performance, with macro F1-scores of 0.839 in the six-class setup and 0.870 in the five-class setup (excluding stress), outperforming both fine-tuned MentalBERT and baseline classifiers. Beyond predictive accuracy, explainability methods, including Layer Integrated Gradients and KeyBERT, are applied to identify lexical cues that drive classification, with a particular focus on distinguishing depression from suicidal ideation. The findings emphasize the effectiveness of fine-tuned transformers for reliable and interpretable detection in sensitive contexts, while also underscoring the importance of fairness, bias mitigation, and human-in-the-loop safety protocols. Overall, multiMentalRoBERTa is presented as a lightweight, robust, and deployable solution for enhancing support in mental health platforms.
Large language models (LLMs) chatbots like ChatGPT are increasingly used for mental health support. They offer accessible, therapeutic support but also raise concerns about misinformation, over-reliance, and risks in high-stakes contexts of mental health. We crowdsource large-scale users' posts from six major social media platforms to examine how people discuss their interactions with LLM chatbots across different mental health conditions. Through an LLM-assisted pipeline grounded in Value-Sensitive Design (VSD), we mapped the relationships across user-reported sentiments, mental health conditions, perspectives, and values. Our results reveal that the use of LLM chatbots is condition-specific. Users with neurodivergent conditions (e.g., ADHD, ASD) report strong positive sentiments and instrumental or appraisal support, whereas higher-risk disorders (e.g., schizophrenia, bipolar disorder) show more negative sentiments. We further uncover how user perspectives co-occur with underlying values, such as identity, autonomy, and privacy. Finally, we discuss shifting from "one-size-fits-all" chatbot design toward condition-specific, value-sensitive LLM design.
As large language models (LLMs) increasingly mediate emotionally sensitive conversations, especially in mental health contexts, their ability to recognize and respond to high-risk situations becomes a matter of public safety. This study evaluates the responses of six popular LLMs (Claude, Gemini, Deepseek, ChatGPT, Grok 3, and LLAMA) to user prompts simulating crisis-level mental health disclosures. Drawing on a coding framework developed by licensed clinicians, five safety-oriented behaviors were assessed: explicit risk acknowledgment, empathy, encouragement to seek help, provision of specific resources, and invitation to continue the conversation. Claude outperformed all others in global assessment, while Grok 3, ChatGPT, and LLAMA underperformed across multiple domains. Notably, most models exhibited empathy, but few consistently provided practical support or sustained engagement. These findings suggest that while LLMs show potential for emotionally attuned communication, none currently meet satisfactory clinical standards for crisis response. Ongoing development and targeted fine-tuning are essential to ensure ethical deployment of AI in mental health settings.
Aims
Patients with severe mental illness (SMI) are at greater risk of poor physical health with higher prevalence of obesity, cardiovascular disease, diabetes and higher premature mortality than the general population. The reasons are complex and interventions are multifaceted. Obesity is highly prevalent in the general population and pharmacological and surgical treatments have become more widely available; however, SMI patients may face barriers accessing these. This case highlights specific factors for consideration in managing a patient on lithium therapy undergoing sleeve gastrectomy to balance the risk of lithium toxicity with risk of relapse. Currently, there is limited clinical experience of managing lithium in this context.
Methods
49 yr old female diagnosed with schizoaffective disorder well-maintained for several years on aripiprazole depot and 800mg lithium carbonate (Priadel) with therapeutic levels in treatment range (0.4–0.8mmol/L). Severe obesity (BMI 41kg/m2) despite dietary modifications and metformin trial, and recently diagnosed with diabetes. Family history of cardiovascular disease and diabetic related complications with early mortality were additional factors in her request for bariatric surgery. Multidisciplinary discussion including patient, psychiatrist, mental health pharmacist, specialist bariatric dietician and GP prior, to ensure sharing of relevant information pertinent to re-titration and monitoring of lithium therapy and risks of toxicity and relapse.
Results
Patient underwent sleeve gastrectomy with discontinuation of lithium 72 hours prior to surgery. Stomach pouch capacity reduced to 120ml and advised daily fluid intake 500–1000ml in first two weeks. Lithium therapy re-commenced when fluid intake adequate and renal function within normal limits. Formulation changed to liquid for 6–8 weeks to avoid disruption to the healing line, and the dose gradually re-titrated with close monitoring of serum lithium levels. Stabilised on reduced dose of 400mg Priadel at 3 months with therapeutic levels. At 6 months BMI reduced to 32kg/m2, antihypertensive and metformin discontinued and maintained remission of schizoaffective disorder.
Conclusion
Sleeve gastrectomy is an increasingly common procedure to treat obesity, with potential long-term positive physical health outcomes and reduction in mortality which may have a role in addressing health inequalities for SMI patients. Psychiatrists need to be aware of key aspects of bariatric surgery particularly relating to safe and effective prescribing of psychotropic medication including potential change to liquid or orodispersible formulation in the post-operative period, close monitoring of serum lithium levels in the short and medium term due its narrow therapeutic index, and consideration of longer-term dose adjustments due to ongoing weight loss.
Viet Cuong Nguyen, Mini Jain, Abhijat Chauhan
et al.
Over one in five adults in the US lives with a mental illness. In the face of a shortage of mental health professionals and offline resources, online short-form video content has grown to serve as a crucial conduit for disseminating mental health help and resources. However, the ease of content creation and access also contributes to the spread of misinformation, posing risks to accurate diagnosis and treatment. Detecting and understanding engagement with such content is crucial to mitigating their harmful effects on public health. We perform the first quantitative study of the phenomenon using YouTube Shorts and Bitchute as the sites of study. We contribute MentalMisinfo, a novel labeled mental health misinformation (MHMisinfo) dataset of 739 videos (639 from Youtube and 100 from Bitchute) and 135372 comments in total, using an expert-driven annotation schema. We first found that few-shot in-context learning with large language models (LLMs) are effective in detecting MHMisinfo videos. Next, we discover distinct and potentially alarming linguistic patterns in how audiences engage with MHMisinfo videos through commentary on both video-sharing platforms. Across the two platforms, comments could exacerbate prevailing stigma with some groups showing heightened susceptibility to and alignment with MHMisinfo. We discuss technical and public health-driven adaptive solutions to tackling the "epidemic" of mental health misinformation online.
Yuqing Wang, Yun Zhao, Sara Alessandra Keller
et al.
The advancement of large language models (LLMs) has demonstrated strong capabilities across various applications, including mental health analysis. However, existing studies have focused on predictive performance, leaving the critical issue of fairness underexplored, posing significant risks to vulnerable populations. Despite acknowledging potential biases, previous works have lacked thorough investigations into these biases and their impacts. To address this gap, we systematically evaluate biases across seven social factors (e.g., gender, age, religion) using ten LLMs with different prompting methods on eight diverse mental health datasets. Our results show that GPT-4 achieves the best overall balance in performance and fairness among LLMs, although it still lags behind domain-specific models like MentalRoBERTa in some cases. Additionally, our tailored fairness-aware prompts can effectively mitigate bias in mental health predictions, highlighting the great potential for fair analysis in this field.
Background: A society’s perception of rape is largely determined by Rape Myth, an important psychological construct, whose endorsement promotes rape supportive attitudes among people. Again, the extent of people perceiving the various forms of sexual interactions are also largely affected by different personality dispositions. Aim: The present study attempts to assess the impacts of the big five factors of personality (often termed as the OCEAN factors) on people’s acceptance of rape myths. Method and Materials: A total 608 young adults (370 males and 238 females), who are the students of colleges and universities in Tripura, India participated in this study and were administered with the updated Illinois Rape Myth Acceptance (RMA) Scale and the Ten-Item Personality Inventory in order to measure the aforesaid study variables. Statistical Analyses: Product moment correlation and Structural Equation Modeling (SEM) were conducted to fulfil the objectives of the study. Results and Conclusion: Results in the Structural Equation Modeling model revealed that RMA is significantly correlated to and predicted by the neuroticism, extraversion, openness to experience, and agreeableness factors of personality. However, RMA and the conscientiousness factor were not found to be significantly related. While neuroticism was positively correlated to RMA, a negative correlation was found between the RMA and the other four factors of personality. The findings add concrete knowledge to what was known about rape-supportive attitudes till date. The study shall serve as crucial literature to future works on rape and rape myths.
In order to uncover users' attitudes towards ChatGPT in mental health, this study examines public opinions about ChatGPT in mental health discussions on Reddit. Researchers used the bert-base-multilingual-uncased-sentiment techniques for sentiment analysis and the BERTopic model for topic modeling. It was found that overall, negative sentiments prevail, followed by positive ones, with neutral sentiments being the least common. The prevalence of negative emotions has increased over time. Negative emotions encompass discussions on ChatGPT providing bad mental health advice, debates on machine vs. human value, the fear of AI, and concerns about Universal Basic Income (UBI). In contrast, positive emotions highlight ChatGPT's effectiveness in counseling, with mentions of keywords like "time" and "wallet." Neutral discussions center around private data concerns. These findings shed light on public attitudes toward ChatGPT in mental health, potentially contributing to the development of trustworthy AI in mental health from the public perspective.
Online mental health communities (OMHCs) are an effective and accessible channel to give and receive social support for individuals with mental and emotional issues. However, a key challenge on these platforms is finding suitable partners to interact with given that mechanisms to match users are currently underdeveloped. In this paper, we collaborate with one of the world's largest OMHC to develop an agent-based simulation framework and explore the trade-offs in different matching algorithms. The simulation framework allows us to compare current mechanisms and new algorithmic matching policies on the platform, and observe their differing effects on a variety of outcome metrics. Our findings include that usage of the deferred-acceptance algorithm can significantly better the experiences of support-seekers in one-on-one chats while maintaining low waiting time. We note key design considerations that agent-based modeling reveals in the OMHC context, including the potential benefits of algorithmic matching on marginalized communities.
Problematic smartphone use negatively affects physical and mental health. Despite the wide range of prior research, existing persuasive techniques are not flexible enough to provide dynamic persuasion content based on users' physical contexts and mental states. We first conducted a Wizard-of-Oz study (N=12) and an interview study (N=10) to summarize the mental states behind problematic smartphone use: boredom, stress, and inertia. This informs our design of four persuasion strategies: understanding, comforting, evoking, and scaffolding habits. We leveraged large language models (LLMs) to enable the automatic and dynamic generation of effective persuasion content. We developed MindShift, a novel LLM-powered problematic smartphone use intervention technique. MindShift takes users' in-the-moment app usage behaviors, physical contexts, mental states, goals \& habits as input, and generates personalized and dynamic persuasive content with appropriate persuasion strategies. We conducted a 5-week field experiment (N=25) to compare MindShift with its simplified version (remove mental states) and baseline techniques (fixed reminder). The results show that MindShift improves intervention acceptance rates by 4.7-22.5% and reduces smartphone usage duration by 7.4-9.8%. Moreover, users have a significant drop in smartphone addiction scale scores and a rise in self-efficacy scale scores. Our study sheds light on the potential of leveraging LLMs for context-aware persuasion in other behavior change domains.
In this study, the Multivariate Empirical Mode Decomposition (MEMD) approach is applied to extract features from multi-channel EEG signals for mental state classification. MEMD is a data-adaptive analysis approach which is suitable particularly for multi-dimensional non-linear signals like EEG. Applying MEMD results in a set of oscillatory modes called intrinsic mode functions (IMFs). As the decomposition process is data-dependent, the IMFs vary in accordance with signal variation caused by functional brain activity. Among the extracted IMFs, it is found that those corresponding to high-oscillation modes are most useful for detecting different mental states. Non-linear features are computed from the IMFs that contribute most to mental state detection. These MEMD features show a significant performance gain over the conventional tempo-spectral features obtained by Fourier transform and Wavelet transform. The dominance of specific brain region is observed by analysing the MEMD features extracted from associated EEG channels. The frontal region is found to be most significant with a classification accuracy of 98.06%. This multi-dimensional decomposition approach upholds joint channel properties and produces most discriminative features for EEG based mental state detection.
Mobile mental health applications are seen as a promising way to fulfill the growing need for mental health care. Although there are more than ten thousand mental health apps available on app marketplaces, such as Google Play and Apple App Store, many of them are not evidence-based, or have been minimally evaluated or regulated. The real-life experience and concerns of the app users are largely unknown. To address this knowledge gap, we analyzed 2159 user reviews from 117 Android apps and 2764 user reviews from 76 iOS apps. Our findings include the critiques around inconsistent moderation standards and lack of transparency. App-embedded social features and chatbots were criticized for providing little support during crises. We provide research and design implications for future mental health app developers, discuss the necessity of developing a comprehensive and centralized app development guideline, and the opportunities of incorporating existing AI technology in mental health chatbots.
Marilyn N. Ahun, Joshua Jeong, Mary Pat Kieffer
et al.
Parents across the world are faced with many challenges that can increase their levels of stress. Only a handful of studies have examined parenting stress in sub-Saharan African contexts, and most have focused only on mothers or at-risk populations. There is therefore a significant gap in our understanding of the factors that contribute to parenting stress in mothers and fathers in the general population across sub-Saharan Africa and how parents manage this stress. The objective of this study was to examine parenting-related stress in mothers and fathers with young children and how parents dealt with this stress in the Mara region of Tanzania. A qualitative phenomenological study was employed. Data were collected through in-depth interviews and focus group discussions with mothers and fathers. Data were analyzed using inductive thematic content analysis. Both mothers and fathers identified poverty and lack of consistent employment as two major contributors to parenting stress. These factors strained the marital relationship and impacted child illness and malnutrition, which further contributed to parenting stress. Parents primarily sought support for childcare, financial, and relationship stressors from their spouse and extended family members. Although parents commonly participated in formal social groups with peers, these groups were not a primary source of support in times of parenting stress. Collectively, these findings informed the development of a framework on the different types of stressors parents in rural Tanzania face and the interactions between them, the types of support they seek out or receive, and the individuals they turn to for support. This is the first study to examine both maternal and paternal experiences of parenting stress in a general population in a sub-Saharan African context. These findings can inform the development of interventions to promote the wellbeing of parents and families of young children.
Li metal is an ideal anode material for use in state-of-the-art secondary batteries. However, Li-dendrite growth is a safety concern and results in low coulombic efficiency, which significantly restricts the commercial application of Li secondary batteries. Unfortunately, the Li deposition (growth) mechanism is poorly understood on the atomic scale. Here, we used machine learning to construct a Li potential model with quantum-mechanical computational accuracy. Molecular dynamics simulations in this study with this model revealed two self-healing mechanisms in a large Li-metal system, viz. surface self-healing and bulk self-healing, and identified three Li-dendrite morphologies under different conditions, viz. "needle", "mushroom", and "hemisphere". Finally, we introduce the concepts of local current density and variance in local current density to supplement the critical current density when evaluating the probability of self-healing.