Student well-being, encompassing mental, social, cognitive, and behavioral domains, is increasingly compromised by academic stress, social isolation, and sedentary lifestyles. Horticulture-based interventions (HBIs), involving plant-based activities, have shown potential in promoting holistic health across populations. Nevertheless, no systematic review has synthesized global evidence for its effects on students. This systematic review aimed to evaluate HBI’s impact on students’ well-being, synthesizing global evidence to inform educational and therapeutic practices. This systematic review was registered in PROSPERO (CRD420251250712). Following PRISMA guidelines, we searched PubMed, Web of Science, MEDLINE, EMBASE, and APA PsycInfo from inception to 30 June 2025. Keywords were used to search for related articles. Fifteen studies (n > 2000 students, aged 5–18 years) from South Korea (n = 8), Taiwan (n = 3), Chinese Mainland (n = 1), Hong Kong, China (n = 1), Italy (n = 1), and the United States (n = 1) were included for analysis. Results showed that HBI has the potential to enhance emotional/psychological well-being (e.g., stress reduction, mood improvement), social well-being (e.g., peer relations, social skills), cognitive and education benefits (e.g., attention, academic attitudes), and physical and overall health benefits (e.g., physical activity, quality of life). HBI may contribute to multifaceted student well-being, particularly emotional and social domains. This systematic review provides a reference for educators to integrate horticultural programs into the curriculum. Government and school policies may consider funding school gardens. Future randomized controlled trials with diverse populations are needed to address limitations like small samples and geographic bias.

BACKGROUND:Virtual humans are embodied conversational agents that are starting to be investigated in healthcare to combat critical staff shortages. It is important to investigate their ability to deliver empathic consultations compared to human physicians. Incorporating good communication behaviours may help to build trust and adherence. AIM:This study's primary aim was to investigate the effects of clinical communication skills (good or poor) and physician type (virtual or human physician) on participants' perceptions of a brief medical consultation. DESIGN:A 2 (a virtual or human physician) x 2 (good or poor communication skills) parallel randomised controlled trial. SETTING:Community-dwelling adults recruited over social media. PARTICIPANTS:One hundred twenty-four adults aged 18 years or over. INTERVENTIONS:Participants watched a video of an actor in a medical consultation with either a virtual physician or a human physician, with either good or poor communication skills. MAIN OUTCOME MEASURES:Ratings of physician empathy, warmth, trust, competence, and adherence intention. RESULTS:Consultations with good communication skills were rated better on physician empathy, warmth, trust, competence, and adherence intention for both physician types (P < .005). The virtual physician was rated more empathetic than the human physician in the poor communication skill condition (P < .001). In addition, the actor was rated less likely to adhere to the virtual physician than the human physician (P = .010). CONCLUSIONS:Building good clinical communication skills into virtual physicians can encourage warm, trusting and competent relationships, which may improve adherence. Future research could investigate dynamic and real-world patient settings. Key messages What is already known on this topic: Good communication skills can improve patient trust and adherence in medical consultations. It is not known whether communication skills affect perceptions of virtual physicians compared to human physicians. What this study adds: Good clinical communication skills improve perceptions of physician empathy, warmth, trust, competence, and adherence intention, in virtual physicians similar to human physicians. Perceptions may favour virtual physicians compared to human physicians when they have poor communication skills. How this study might affect research, practice or policy: The study supports the use of virtual physicians with good communication skills. Three research questions-as snappy bullet points-that outline the current research questions-i.e. not the ones that you have answered but the ones that remain or have emerged as a result of your work. Will patients trust virtual physicians in real clinical consultations? Will trust in virtual physicians depend on the severity of the patient's condition? Will adherence differ between physicians and virtual physicians after real patient consultations?

Highly realistic embodied conversational agents (virtual humans) are starting to be used in healthcare, yet we know little about how their human likeness affects self-disclosure compared to other digital assessment methods. This research aimed to investigate the effects of a computer's human likeness on disclosure during a psychosocial assessment. 160 participants (mean age 24 years, 117 females, 42 males, 1 gender diverse) were randomized to receive a psychosocial interview from a realistic virtual human, a text chatbot, or an online questionnaire. The assessment comprised 18 close-ended items and six open-ended questions on diet, exercise, sexual practices, substance use, recent emotional experiences and loneliness. Socially desirable responding and unwillingness to respond were identified from answers, amount of disclosure was assessed from word count, and perceived anthropomorphism was assessed using self-report. Results demonstrated that for sensitive questions, there was higher socially desirable responding in the virtual human group as shown by significantly lower loneliness scores and higher rates of declining to answer a question compared to the other groups. For non-sensitive questions, socially desirable responding and proportion of declined answers did not differ by group. Participants in the virtual human group used more words when reporting stressful events and positive emotional experiences. Thematic analysis showed people felt rapport with both the virtual human and chatbot, but some felt social evaluative pressure with the virtual human. This supports theories that as human likeness cues increase, humans treat robots more socially. These findings should be considered when humanlike technologies are used clinically.

This study investigated the acceptability of a hospital robot’s responses to hypothetical patients, considering empathetic and non-empathetic conditions. It also explored the benefits and challenges of using social robots in hospitals, and areas for improving empathy. Using a mixed-methods survey, participants reported their perceptions of acceptability, usefulness, satisfaction, and empathy after watching 8 videos of a robot greeting, checking COVID-19 index symptoms, entertaining and giving directions to a patient actor. The robot was empathetic in 4 videos and non-empathetic in 4 other videos. Open-ended questions explored the benefits and challenges of using hospital robots, and ideas for improvement. Descriptive statistics and paired samples Wilcoxon-signed rank tests were conducted on quantitative data. Qualitative data were analyzed using an inductive content synthesis approach. Participants (N = 147) were most comfortable with the robot giving directions (mean: 4.34; SD: 0.90) and least comfortable with checking COVID-19 symptoms (mean: 3.06; SD: 1.21). Participants were significantly more satisfied with the empathetic robot and rated it as more empathetic, useful, and having more natural speech and gestures (p’s < 0.001). They were less likely to prefer a human when the empathetic robot was performing the tasks (p = 0.005). Benefits were related to the healthcare system, staff and patients, whereas the technology, the user experience, the robot not being human, and logistics could create challenges. Empathetic robots may be perceived as more acceptable. Future work should explore how to create more genuine interactions, identify low-risk areas for automation in healthcare and focus on synchronizing responses.

Conversational agents (CAs) are revolutionizing human-computer interaction by evolving from text-based chatbots to empathetic digital humans (DHs) capable of rich emotional expressions. This paper explores the integration of neural and physiological signals into the perception module of CAs to enhance empathetic interactions. By leveraging these cues, the study aims to detect emotions in real-time and generate empathetic responses and expressions. We conducted a user study where participants engaged in conversations with a DH about emotional topics. The DH responded and displayed expressions by mirroring detected emotions in real-time using neural and physiological cues. The results indicate that participants experienced stronger emotions and greater engagement during interactions with the Empathetic DH, demonstrating the effectiveness of incorporating neural and physiological signals for real-time emotion recognition. However, several challenges were identified, including recognition accuracy, emotional transition speeds, individual personality effects, and limitations in voice tone modulation. Addressing these challenges is crucial for further refining Empathetic DHs and fostering meaningful connections between humans and artificial entities. Overall, this research advances human-agent interaction and highlights the potential of real-time neural and physiological emotion recognition in creating empathetic DHs.

Objective: Relaxation delivered via audiotapes can reduce stress and improve wound healing. Virtual humans are a promising technology to deliver relaxation, but robust research is needed into their effectiveness. This randomised controlled trial investigated whether relaxation delivered by a virtual human could improve healing and reduce stress after an experimental wound. Methods: A total of 159 healthy adults underwent a tape-stripping wounding procedure and were randomly assigned to relaxation delivered by a virtual human, human audiotape, or a control condition. Skin barrier recovery (SBR) was measured by assessing changes in transepidermal water loss at baseline, post-tape-stripping, and post-intervention. Psychological and physiological variables were measured over the session. Participants’ perceptions of the interventions were assessed. Results: There were no significant differences in SBR between conditions. All conditions experienced significant improvements in the psychological variables, heart rate, and cortisol over time. After controlling for the baseline values, the virtual human and audiotape conditions were significantly more relaxed post-intervention than the control condition (p = 0.005), the audiotape condition had lower post-intervention anxiety than the control condition (p = 0.016), and alpha-amylase was significantly reduced in the virtual human group compared with the audiotape (p = 0.041). The audiotape received the highest satisfaction and engagement ratings, with qualitative results suggesting the appearance and lip-syncing of the virtual human could be improved. Conclusions: Relaxation instructions delivered by a virtual human increased participants’ relaxation levels with similar effects to traditional audiotapes. Furthermore, it reduced physiological stress indices. Further work with other wound types and stressed samples is needed. The voice and interactiveness of the virtual human should be improved to promote greater engagement and satisfaction.

Background:Artificial intelligence (AI) chatbots have shown competency in a range of areas, including clinical note taking, diagnosis, research, and emotional support. An obesity epidemic, alongside a growth in novel injectable pharmacological solutions, has put a strain on limited resources. Objective:This study aimed to investigate the use of a chatbot integrated with a digital avatar to create a "digital clinician." This was used to provide mandatory patient education for those beginning semaglutide once-weekly self-administered injections for the treatment of overweight and obesity at a national center. Methods:A "digital clinician" with facial and vocal recognition technology was generated with a bespoke 10- to 15-minute clinician-validated tutorial. A feasibility randomized controlled noninferiority trial compared knowledge test scores, self-efficacy, consultation satisfaction, and trust levels between those using the AI-powered clinician avatar onsite and those receiving conventional semaglutide education from nursing staff. Attitudes were recorded immediately after the intervention and again at 2 weeks after the education session. Results:A total of 43 participants were recruited, 27 to the intervention group and 16 to the control group. Patients in the "digital clinician" group were significantly more knowledgeable postconsultation (median 10, IQR 10-11 vs median 8, IQR 7-9.3; P<.001). Patients in the control group were more satisfied with their consultation (median 7, IQR 6-7 vs median 7, IQR 7-7; P<.001) and had more trust in their education provider (median 7, IQR 4.8-7 vs median 7, IQR 7-7; P<.001). There was no significant difference in reported levels of self-efficacy (P=.57). 81% (22/27) participants in the intervention group said they would use the resource in their own time. Conclusions:Bespoke AI chatbots integrated with digital avatars to create a "digital clinician" may perform health care education in a clinical environment. They can ensure higher levels of knowledge transfer yet are not as trusted as their human counterparts. "Digital clinicians" may have the potential to aid the redistribution of resources, alleviating pressure on bariatric services and health care systems, the extent to which remains to be determined in future studies.

ObjectiveCognitive behavioural therapy (CBT) has shown efficacy in improving mental health and symptoms of functional dyspepsia (FD), a prevalent disorder of gut-brain interaction (DGBI). However, FD-specific CBT is not widely available or scalable. Therefore, this study explored the perspectives of patients with FD and clinicians who treat them on the use of digital CBT-based interventions.Methods and measuresThis qualitative study involved semi-structured interviews with 21 patients with FD and 10 clinicians. Iterative, inductive thematic analysis was conducted.ResultsThree patient themes were developed: (1) their experiences contributing to coping difficulties, emphasising the complexity of self-management; (2) a desire for more personalised options, highlighting the need for accessible, FD-tailored CBT; and (3) concerns regarding human contact, safety, usability, and data security. Four clinician themes were also developed: (1) digital CBT bridges a gap in psychological support within gastroenterology; (2) perceived clinical utility of CBT; (3) digital CBT allows patient self-management; and (4) clinical concerns, including symptom tracking and patient safety.ConclusionBoth patients and clinicians recognised the value and utility of digital CBT for FD. Tailored, digital CBT could inform and improve current management, making FD-specific psychological support more accessible.

Background:Rising global obesity rates demand effective weight management strategies from general practitioners (GPs). However, time constraints, training gaps, and low confidence often impede GPs' ability to conduct weight-based conversations. This pilot study assessed the feasibility and preliminary effectiveness of an AI-driven Virtual Human (VH) obesity education and communication-skills training tool, specifically designed to address these challenges and enhance obesity education and communication-skills among GPs. Methods:A pilot feasibility study with a pre-post survey design evaluated the impact of the VH tool on knowledge, self-efficacy, empathy toward patients with obesity, and confidence in clinical consultations. Participant perceptions, trust, and intention to use the VH tool were explored. Paired-sample t-tests were conducted to evaluate within-group mean differences. Descriptive statistics were used to evaluate feasibility and acceptability. Results:A total of 22 GPs were recruited. Despite some attrition, significant improvements were observed in knowledge (p = 0.006), self-efficacy (p = 0.001), and combined empathy and confidence scores (p = 0.002). Alongside these improvements, participants demonstrated positive perceptions of the tool, high trust in the VH, and a strong intention to implement the learned strategies. Conclusions:This pilot study demonstrates the potential of an AI-driven VH tool to enhance GP obesity education and communication skills. The observed improvements in key outcomes support the potential of VH technology in medical education on obesity. To further establish the efficacy and explore the broader applicability, future research should focus on larger, controlled trials across various provider groups. Overall, these preliminary observations highlight a promising avenue for enhancing the skills of a wider range of providers in the obesity treatment space.

Social robots are increasingly being applied in hospitals assisting with tasks that conventionally require interpersonal interactions. An understanding of how social robots engage with users is crucial as this will inform any strength or weakness associated with the current engagement approaches. Prior reviews on healthcare social robots have mainly focused on summarising the types of services performed by the robots. We conducted a scoping review examining the dynamics of human-robot interactions (HRI) within a hospital setting. In particular, we provided an overview on the service contexts of the HRI, robot engagement behaviours, and interaction outcomes. The scoping review was conducted following the Arksey and O’Malley 5-step method and was reported using the PRISMA-ScR extension for scoping reviews. The analysis of 26 studies found that social robots delivered a wide range of hospital services to various users, from greeting and educating visitors to social companionship, supporting healthcare delivery, carrying goods, and training staff. The robots interacted with users using verbal and non-verbal behaviours to convey intentions and emotions. Personalised behaviours were adopted by some robots to tailor to the user needs and preferences. The HRI was mostly positive with high user satisfaction, good user engagement, and improved health outcomes. Frustration and fear towards the interactions were also reported by some users. Optimising multimodal and personalised engagement are potential avenues for achieving effective and warm HRI. Further focus on understanding the contexts of HRI could help support the implementation of social robots in hospital settings.

Mild cognitive impairment (MCI) is an early stage of cognitive decline that significantly increases the risk of dementia, making early interventions crucial for maintaining cognitive health in older adults. Our research takes a co-design approach to understand, design, and evaluate how socially assistive robots and a virtual human can promote lifestyle changes for people with MCI (pwMCI), potentially improving their cognitive health. Through an iterative refinement process, we aim to develop adaptable, user-friendly, and sustainable technologies that foster long-term engagement across diverse settings. Ultimately, our goal is to improve cognitive health, quality of life, emotional well-being, and loneliness in pwMCI.

Research has shown that the composition of breath can differ based on the human's behavioral patterns and mental and physical states immediately before being collected. These breath-collection techniques have also been extended to observe the general processes occurring in groups of humans and can link them to what those groups are collectively experiencing. In this research, we applied machine learning techniques to the breath data collected from cinema audiences. These techniques included XGBOOST Regression, Hierarchical Clustering, and Item Basket analyses created using the Apriori algorithm. They were conducted to find associations between the biomarkers in the crowd's breath and the movie's audio-visual stimuli and thematic events. This analysis enabled us to directly link what the group was experiencing and their biological response to that experience. We first extracted visual and auditory features from a movie to achieve this. We compared it to the biomarkers in the crowd's breath using regression and pattern mining techniques. Our results supported the theory that a crowd's collective experience directly correlates to the biomarkers in the crowd's breath. Consequently, these findings suggest that visual and auditory experiences have predictable effects on the human body that can be monitored without requiring expensive or invasive neuroimaging techniques.

Background: Evidence suggests that countries with higher Covid-19 infection rates experienced poorer mental health. This study examined whether hair cortisol reduced over time in New Zealand, a country that managed to eliminate the virus in the first year of the pandemic due to an initial strict lockdown. Methods: A longitudinal cohort study assessed self-reported stress, anxiety and depression and collected hair samples that were analyzed for cortisol, across two waves in 2020. The sample consisted of 44 adults who each returned two 3 cm hair samples and completed self-reports. Hair cortisol was assessed per centimetre. Results: Hair cortisol reduced over time (F (5, 99.126) = 10.15, p < .001, partial eta squared = 0.19), as did anxiety and depression. Higher hair cortisol was significantly associated with more negative life events reported at wave two (r = 0.30 segment 1, r = 0.34 segment 2, p < .05), but not anxiety or depression. Conclusions: Strict virus control measures may not only reduce infection rates, but also reduce psychological distress, and hair cortisol over time.

Aims: Users’ empathy towards artificial agents can be influenced by the agent’s expression of emotion. To date, most studies have used a Wizard of Oz design or manually programmed agents’ expressions. This study investigated whether autonomously animated emotional expression and neural voices could increase user empathy towards a Virtual Human. Methods: 158 adults participated in an online experiment, where they watched videos of six emotional stories generated by ChatGPT. For each story, participants were randomly assigned to a virtual human (VH) called Carina, telling the story with either (1) autonomous expressive or non-expressive animation, and (2) a neural or standard text-to-speech (TTS) voice. After each story, participants rated how well the animation and voice matched the story, and their cognitive, affective, and subjective empathy towards Carina were evaluated. Qualitative data were collected on how well participants thought Carina expressed emotion. Results: Autonomous emotional expression enhanced the alignment between the animation and voice, and improved subjective, cognitive, and affective empathy. The standard voice was rated as matching the fear and sad stories better, the sad animation was better, creating greater subjective and cognitive empathy for the sad story. Trait empathy and ratings of how well the animation and voice matched the story predicted subjective empathy. Qualitative analysis revealed that the animation conveyed emotions more effectively than the voice, and emotional expression was associated with increased empathy. Conclusion: Autonomous emotional animation of VHs can improve empathy towards AI-generated stories. Further research is needed on voices that can dynamically change to express different emotions.

Objective Cardiac inherited diseases can have considerable psychosocial effects, including lifestyle limitations, anxiety and depression. Most research to date on patient experiences of CID has been conducted with people from Western cultures, yet culture can shape patient views and experiences of health. The aim of this research was to explore the experiences and perspectives of Māori and Pasifika living with a cardiac inherited disease (CID).Methods and Measures Semi-structured interviews were conducted with 14 Māori and 14 Pasifika patients living with a cardiac inherited disease and seven of their family members, using Talanoa and Kaupapa Māori methodologies. Themes from the interviews were identified using interpretative phenomenological analysis.Results Three common themes were identified as important in shaping participants' perceptions and experiences of CID: (1) difficulty in understanding the disease as separate from symptoms, (2) considering ancestors and future generations and (3) the role of spirituality and religion.Conclusion This study highlights a gap between indigenous patients' understanding of CID and the western biomedical approach. Patients' understanding and treatment behaviours depend on symptoms, familial ties and spirituality. The findings support the need for transparency and culturally appropriate practices in healthcare. Considering these aspects may help to reduce health inequities for these populations.

Remote communication has become pervasive, yet its impact on human emotions, empathy, and physiological responses remains unclear. This study addresses this gap by investigating how emotional video-mediated conversations differ from face-to-face interactions, focusing on behavioral and physiological responses. We create a multimodal dataset of Electrodermal Activity (EDA) signals, Photoplethysmography (PPG) signals, and facial videos from two people conversing about emotional topics in face-to-face and remote video-mediated conditions. We use a series of repeated measures ANOVA with aligned rank transform to compare face-to-face to remote conversation in terms of heart rate activity, electrodermal activity, and facial action units. We also explore how subjective empathy between people varies in these two conditions. Our findings reveal significant differences in physiological responses between face-to-face and remote conversation and variations in perceived empathy based on interaction setting, highlighting the nuanced influence of communication channels. We also show that we can recognize emotions more accurately when we pre-train a random forest classifier with one condition’s data (an increase of 20% to 45% for various modalities). Finally, we discuss the research findings and limitations and offer insights for optimizing human–computer interaction and understanding human emotional responses in an increasingly tech-mediated world.