Explore the vast range of titles available.

Artificial intelligence (AI) is increasingly being applied in various health care services due to its enhanced efficiency and accuracy. As the population ages, AI-based health technologies could be a potent tool in older adults' health care to address growing, complex, and challenging health needs. This study aimed to investigate perspectives on and acceptability of the use of AI-led health technologies among older adults and the potential challenges that they face in adopting them. The findings from this inquiry could inform the designing of more acceptable and user-friendly AI-based health technologies. The objectives of the study were (1) to investigate the attitudes and perceptions of older adults toward the use of AI-based health technologies; (2) to identify potential facilitators, barriers, and challenges influencing older adults' preferences toward AI-based health technologies; and (3) to inform strategies that can promote and facilitate the use of AI-based health technologies among older adults. This study adopted a qualitative descriptive design. A total of 27 community-dwelling older adults were recruited from a local community center. Three sessions of semistructured interviews were conducted, each lasting 1 hour. The sessions covered five key areas: (1) general impressions of AI-based health technologies; (2) previous experiences with AI-based health technologies; (3) perceptions and attitudes toward AI-based health technologies; (4) anticipated difficulties in using AI-based health technologies and underlying reasons; and (5) willingness, preferences, and motivations for accepting AI-based health technologies. Thematic analysis was applied for data analysis. The Theoretical Domains Framework and the Capability, Opportunity, Motivation, and Behavior (COM-B) model behavior change wheel were integrated into the analysis. Identified theoretical domains were mapped directly to the COM-B model to determine corresponding strategies for enhancing the acceptability of AI-based health technologies among older adults. The analysis identified 9 of the 14 Theoretical Domains Framework domains-knowledge, skills, social influences, environmental context and resources, beliefs about capabilities, beliefs about consequences, intentions, goals, and emotion. These domains were mapped to 6 components of the COM-B model. While most participants acknowledged the potential benefits of AI-based health technologies, they emphasized the irreplaceable role of human expertise and interaction. Participants expressed concerns about the usability of AI technologies, highlighting the need for user-friendly and tailored AI solutions. Privacy concerns and the importance of robust security measures were also emphasized as critical factors affecting their willingness to adopt AI-based health technologies. Integrating AI as a supportive tool alongside health care providers, rather than regarding it as a replacement, was highlighted as a key strategy for promoting acceptance. Government support and clear guidelines are needed to promote ethical AI implementation in health care. These measures can improve health outcomes in the older adult population by encouraging the adoption of AI-driven health technologies.

Machine learning (ML) and deep learning (DL) show promise for fall risk prediction, but prior reviews focused mainly on real-time fall detection, in-hospital falls, or conventional statistical models. The performance of ML-DL-based models for predicting future falls in community-dwelling older adults remains unclear. This study aimed to review ML-DL studies for predicting future falls among community-dwelling older adults and meta-analyze discrimination where feasible. Six databases were searched from inception to September 23, 2024, with updates on August 31, 2025, and February 28, 2026. We included longitudinal studies developing or validating ML-DL models to predict future falls in community-dwelling adults aged ≥60 years and excluded real-time detection, simulated or no fall, and inpatient studies. Risk of bias was assessed using the Prediction Model Risk of Bias Assessment Tool (PROBAST). Areas under the curve (AUCs) were meta-analyzed using Hartung-Knapp-Sidik-Jonkman random-effects models with 95% CIs. Heterogeneity, 95% prediction intervals (PIs), sensitivity analyses, and subgroup analyses were conducted. After screening 10,253 records, 28 (0.3%) studies were included; 18 (64.3%) focused on general older adults. Prediction horizons ranged from 3 months to 7 years, and fall incidence ranged from 1.6% to 46.6%. Twenty-three (82.1%) studies applied ML, and 5 (17.9%) studies used DL. Input modalities included text (n=18, 64.3%), sensor (n=5, 17.9%), image (n=1, 3.6%), and multimodal data (n=4, 14.3%). Common predictors included age, sex, fall history, depression, and basic daily activities. Only one model underwent external validation. Calibration reporting was sparse. All models were rated at high risk of bias. Ten models were meta-analyzed, yielding a pooled AUC of 0.79 (95% CI 0.69-0.87) with extreme heterogeneity (τ2=0.64; τ=0.80; I2=99.8%; Q=4128.99). The confidence-distribution bootstrap PI was 0.20 to 0.99, indicating substantial uncertainty in expected performance across new populations. Subgroup analyses indicated moderation by sample size and population type, with higher discrimination in specific populations than in general samples; however, the specific population subgroup included only 2 studies. Although all participants were community dwelling, some cohorts were recruited through clinically enriched pathways rather than general community sampling. ML-DL models show potential for identifying community-dwelling older adults at elevated future fall risk; however, wide PIs, limited external validation, and high risk of bias suggest real-world performance may be optimistic. The pooled AUC should be interpreted as a summary of reported discrimination under study-specific conditions, predominantly from internally validated, high-risk-of-bias models, rather than as a robust estimate of transportable real-world performance. This review extends prior reviews by focusing on community-dwelling settings and by integrating PROBAST, Hartung-Knapp-Sidik-Jonkman meta-analysis, PIs, and modality-specific synthesis to evaluate both discrimination and uncertainty. Findings support the use of ML-DL models for proactive fall prevention while emphasizing the need for validation and context-specific implementation.

This research letter proposes a novel model design leveraging natively multimodal large language models to identify fall risks and generate visualizations of recommended home environmental modifications, aiming to improve the accessibility and impact of personalized fall prevention advice for older adults. Through a pilot rating study, this work demonstrates that multimodal large language models can generate safe and actionable advice to reduce fall risk in lived spaces of older adults, and also generate realistic edits based on original images. While this concept needs further testing and clinical comparison, it highlights a promising avenue for further innovation of fall prevention tactics.

Background:Identifying persons with a high risk of developing osteoporosis and preventing the occurrence of the first fracture is a health care priority. Most existing osteoporosis screening tools have high sensitivity but relatively low specificity.Objective:We aimed to develop an easily accessible and high-performance preclinical risk screening tool for osteoporosis using a machine learning–based method among the Hong Kong Chinese population.Methods:Participants aged 45 years or older were enrolled from 6 clinics in the 3 major districts of Hong Kong. The potential risk factors for osteoporosis were collected through a validated, self-administered questionnaire and then filtered using a machine learning–based method. Bone mineral density was measured with dual-energy x-ray absorptiometry at the clinics; osteoporosis was defined as a t score of −2.5 or lower. We constructed machine learning models, including gradient boosting machines, support vector machines, and naive Bayes, as well as the commonly used logistic regression models, for the prediction of osteoporosis. The best-performing model was chosen as the final tool, named the Preclinical Osteoporosis Screening Tool (POST). Model performance was evaluated by the area under the receiver operating characteristic curve (AUC) and other metrics.Results:Among the 800 participants enrolled in this study, the prevalence of osteoporosis was 10.6% (n=85). The machine learning–based Boruta algorithm identified 15 significantly important predictors from the 113 potential risk factors. Seven variables were further selected based on their accessibility and convenience for daily self-assessment and health care practice, including age, gender, education level, decreased body height, BMI, number of teeth lost, and the intake of vitamin D supplements, to construct the POST. The AUC of the POST was 0.86 and the sensitivity, specificity, and accuracy were all 0.83. The positive predictive value, negative predictive value, and F1-score were 0.41, 0.98, and 0.56, respectively.Conclusions:The machine learning–based POST was conveniently accessible and exhibited accurate discriminative capabilities for the prediction of osteoporosis; it might be useful to guide population-based preclinical screening of osteoporosis and clinical decision-making.

Conversational artificial intelligence (AI) technologies are increasingly positioned as a response to social isolation, loneliness, and unmet psychosocial needs across health and care contexts. Non-embodied AI-based digital companions have attracted growing attention for their potential to support companionship, social interaction, communication, and psychosocial well-being among older adults, including people living with dementia. However, the evidence base remains underexplored. Terminology is inconsistently applied, systems are variably defined, and studies are distributed across disciplinary silos, limiting critical investigation of how these technologies are conceptualized, designed, and evaluated. This study aims to map and critically synthesize the existing literature on non-embodied AI-based digital companions for older adults in aging, health, and care contexts. The review seeks to describe digital companions, examine methodological approaches, and identify research gaps. This protocol follows the Joanna Briggs Institute methodology for scoping reviews and will be reported in accordance with the PRISMA-ScR (Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews guidelines). A comprehensive search will be conducted across multiple electronic databases, including MEDLINE, APA PsycINFO, CINAHL, Scopus, Web of Science, IEEE Xplore, and ACM Digital Library, as well as selected gray literature sources. The search strategy combined terms related to digital companions, conversational AI, aging, and care contexts, including concepts related to companionship, social interaction, communication, loneliness, and psychosocial support., Eligible studies will include empirical studies involving older adults that examine non-embodied AI-based digital companions designed primarily to support companionship, social interaction, communication, or related psychosocial support in aging, health, and care contexts. Two reviewers (YC and MG) will independently conduct study selection and data charting. Data will be synthesized using descriptive statistics and narrative analysis. This protocol outlines a systematic approach to identifying, selecting, and synthesizing the existing evidence on non-embodied AI-based digital companions. A preliminary PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses)-style search flow for the revised database searches identified 2289 records from MEDLINE, CINAHL, and APA PsycINFO. After duplicate removal and eligibility-based removals before screening, 1978 records remained available for title and abstract screening. The full scoping review will summarize study characteristics, populations, contexts, digital companion features, and methodological trends using descriptive tables and narrative synthesis. As of June 2026, the revised database searches had been completed for MEDLINE, CINAHL, and APA PsycINFO, and title and abstract screening was underway. The full scoping review results are expected to be submitted for publication after screening, data charting, and synthesis are completed. This scoping review is expected to clarify conceptual boundaries, map the scope of current research, and identify knowledge gaps related to non-embodied AI-based digital companions in health, aging, and care contexts. The findings will inform future research, design, and implementation of non-embodied AI-based digital companions.

Interventions and care that can evoke positive emotions and reduce apathy or agitation are important for people with dementia. In recent years, socially assistive robots used for better dementia care have been found to be feasible. However, the immediate responses of people with dementia when they are given multiple sensory modalities from socially assistive robots have not yet been sufficiently elucidated. This study aimed to quantitatively examine the immediate emotional responses of people with dementia to stimuli presented by socially assistive robots using facial expression analysis in order to determine whether they elicited positive emotions. This pilot study adopted a single-arm interventional design. Socially assistive robots were presented to nursing home residents in a three-step procedure: (1) the robot was placed in front of participants (visual stimulus), (2) the robot was manipulated to produce sound (visual and auditory stimuli), and (3) participants held the robot in their hands (visual, auditory, and tactile stimuli). Expression intensity values for \"happy,\" \"sad,\" \"angry,\" \"surprised,\" \"scared,\" and \"disgusted\" were calculated continuously using facial expression analysis with FaceReader. Additionally, self-reported feelings were assessed using a 5-point Likert scale. In addition to the comparison between the subjective and objective emotional assessments, expression intensity values were compared across the aforementioned 3 stimuli patterns within each session. Finally, the expression intensity value for \"happy\" was compared between the different types of robots. A total of 29 participants (mean age 88.7, SD 6.2 years; n=27 female; Japanese version of Mini-Mental State Examination mean score 18.2, SD 5.1) were recruited. The expression intensity value for \"happy\" was the largest in both the subjective and objective assessments and increased significantly when all sensory modalities (visual, auditory, and tactile) were presented (median expression intensity 0.21, IQR 0.09-0.35) compared to the other 2 patterns (visual alone: median expression intensity 0.10, IQR 0.03-0.22; P<.001; visual and auditory: median expression intensity 0.10, IQR 0.04-0.23; P<.001). The comparison of different types of robots revealed a significant increase when all stimuli were presented by doll-type and animal-type robots, but not humanoid-type robots. By quantifying the emotional responses of people with dementia, this study highlighted that socially assistive robots may be more effective in eliciting positive emotions when multiple sensory stimuli, including tactile stimuli, are involved. More studies, including randomized controlled trials, are required to further explore the effectiveness of using socially assistive robots in dementia care.

Wound care is an increasing global challenge, with older adults among those most affected. As populations age, the demand for effective and efficient wound care increases. Over the years, various wound assessment and care techniques have been developed, including digital wound care technology (DWCT), which uses innovative artificial intelligence (AI). Many older adults, especially those living in rural and remote areas, face significant barriers in obtaining timely and effective wound care, leading to poorer health outcomes and increased health care costs related to wound care. These challenges underscore the urgent need to implement wound care models that equitably improve access to care and enhance clinical outcomes, particularly for older adults, to promote healthy aging and age-in-place. Based on evidence from the literature and the initial implementation of a DWCT in 2 community health systems in Ontario, this viewpoint paper encourages clinicians and health care leaders to embrace and expand the implementation of an AI-driven DWCT to address inequities in access to high-quality, timely care. The experiences from these implementations indicate that the use of AI can support clinical decision-making and extend access to care for individuals in rural and remote communities in Canada. By leveraging DWCT powered by AI, health care providers can enhance the accuracy and consistency of wound assessments, improve communication, streamline care processes, and more effectively allocate resources, ultimately aiming to reduce disparities in wound care outcomes. 

People with dementia commonly display behavioral and psychological symptoms, which have multiple negative consequences. Artificial intelligence-based technologies (AITs) have the potential to support earlier detection of the behavioral and psychological symptoms of dementia (BPSD). The recent surge of interest in this topic underscores the need to comprehensively examine the existing evidence. This scoping review aimed to identify and summarize the types and uses of AITs currently used for the early detection of BPSD among people diagnosed with the disease. We also examined which health care professionals were involved, nursing involvement and experience, the care settings in which these technologies are used, and the characteristics of the BPSD that were assessed. Our scoping review was conducted in accordance with the Joanna Briggs Institute manual for scoping reviews. Searches were conducted in March 2025 in the following bibliographic databases: MEDLINE ALL Ovid, Embase, APA PsycINFO Ovid, CINAHL EBSCO, Web of Science Core Collection, the Cochrane Library Wiley, and ProQuest Dissertations and Theses A&I. Additional searches were performed using citation tracking strategies and by consulting the Association for Computing Machinery Digital Library. Eligible studies included primary research involving people with dementia and examining the use of AITs for the detection of BPSD in real-world care settings. After screening 3670 articles for eligibility, the review includes 12 studies. The studies retained were conducted between 2012 and 2025 in 5 countries and encompassed a range of care settings. The AITs used were predominantly based on classic machine learning approaches and used information from environmental sensors, wearable devices, and data recording systems. These studies primarily assessed behavioral and physiological parameters and focused specifically on symptoms, such as agitation and aggression. None of the retained studies explored nurses' roles or their specific skills in using these technologies. The use of AITs for managing BPSD represents an emerging field of research offering novel opportunities to enhance their detection in various health care contexts. We recommended that nurses be actively engaged in developing and assessing these technologies. Future research should prioritize investigations into how effective AITs are across diverse populations, whether they can have a long-term impact on managing BPSD, and whether they can improve the quality of life of patients and caregivers.

The two most commonly used methods to identify frailty are the frailty phenotype and the frailty index. However, both methods have limitations in clinical application. In addition, methods for measuring frailty have not yet been standardized. We aimed to develop and validate a classification model for predicting frailty status using vocal biomarkers in community-dwelling older adults, based on voice recordings obtained from the picture description task (PDT). We recruited 127 participants aged 50 years and older and collected clinical information through a short form of the Comprehensive Geriatric Assessment scale. Voice recordings were collected with a tablet device during the Korean version of the PDT, and we preprocessed audio data to remove background noise before feature extraction. Three artificial intelligence (AI) models were developed for identifying frailty status: SpeechAI (using speech data only), DemoAI (using demographic data only), and DemoSpeechAI (combining both data types). Our models were trained and evaluated on the basis of 5-fold cross-validation for 127 participants and compared. The SpeechAI model, using deep learning-based acoustic features, outperformed in terms of accuracy and area under the receiver operating characteristic curve (AUC), 80.4% (95% CI 76.89%-83.91%) and 0.89 (95% CI 0.86-0.92), respectively, while the model using only demographics showed an accuracy of 67.96% (95% CI 67.63%-68.29%) and an AUC of 0.74 (95% CI 0.73-0.75). The SpeechAI model outperformed the model using only demographics significantly in AUC (t4=8.705 [2-sided]; P<.001). The DemoSpeechAI model, which combined demographics with deep learning-based acoustic features, showed superior performance (accuracy 85.6%, 95% CI 80.03%-91.17% and AUC 0.93, 95% CI 0.89-0.97), but there was no significant difference in AUC between the SpeechAI and DemoSpeechAI models (t4=1.057 [2-sided]; P=.35). Compared with models using traditional acoustic features from the openSMILE toolkit, the SpeechAI model demonstrated superior performance (AUC 0.89) over traditional methods (logistic regression: AUC 0.62; decision tree: AUC 0.57; random forest: AUC 0.66). Our findings demonstrate that vocal biomarkers derived from deep learning-based acoustic features can be effectively used to predict frailty status in community-dwelling older adults. The SpeechAI model showed promising accuracy and AUC, outperforming models based solely on demographic data or traditional acoustic features. Furthermore, while the combined DemoSpeechAI model showed slightly improved performance over the SpeechAI model, the difference was not statistically significant. These results suggest that speech-based AI models offer a noninvasive, scalable method for frailty detection, potentially streamlining assessments in clinical and community settings.

Population aging has emerged as a global concern, older adults' ability to access health knowledge and manage their well-being impacts their health outcomes. In the artificial intelligence era, generative artificial intelligence (GenAI) videos hold promise for enhancing geriatric health management. However, their potential and the needs of older adults in using GenAI videos for health-related purposes deserve a more in-depth investigation. This study aims to explore the application potential and multifaceted needs of older adults in using GenAI videos for health information acquisition and management, while providing actionable recommendations for future aging-friendly GenAI video tools. A qualitative approach was adopted. Twenty older adults (aged ≥60 y) with basic digital literacy were recruited from communities for participation in iterative GenAI video workshops. Semistructured interviews were conducted. Thematic analysis, following Braun and Clarke's 6-phase reflexive framework, was used to identify key themes from interview transcripts. Our results have identified a 3-layer hierarchical structure of needs when older adults interact with GenAI videos. The first layer is technical accessibility, which suggests the direct barriers preventing them from operating GenAI tools independently. The second layer is age-friendly design, which depicts the needs of age-friendly design adaptations, which can help older adults to better use GenAI features. The final layer is integration with professional resources, highlighting the need for integrating professional medical information resources with GenAI tools, so that the GenAI videos can be more appropriately used for health management purposes. By triangulating the core needs of older adults, the capabilities of GenAI, and their remaining gaps, we have found that the existing gaps limit the extent to which the health needs of older adults are met, and also inversely restrain the acceptance and use of GenAI videos by older adults. This study explored the potential and gaps of using GenAI videos in older adults' health management. GenAI videos are prominent for self-management of health, but our triangular model reveals coexisting opportunities and challenges. While the core needs drive the development directions of GenAI video technologies, our study suggests that GenAI health videos can benefit from the improvements in technical capabilities, service innovation, localization, and integration of health resources.