Daeun Lee

Background: Suicide assessments have historically privileged verbal report by the patient, despite the fact that nonverbal behaviors of patients and their clinicians may convey important affective and interpersonal information about suicide risk. Recent advances in computational science enable efficient characterization of rich nonverbal data. Objective: This study aimed to use automated coding to test whether facial action and head motion exhibited by young adults and their clinical interviewers during a widely used suicide assessment can identify suicidal participants. Methods: Participants were a diverse sample of 66 young adults (age: mean 21.32, SD 2.11 years) recruited from the community, half of whom engaged in past-year suicidal behavior (ie, suicidal participants) and half of whom had no history of suicidality (ie, nonsuicidal participants). Facial action units, head pose, and eye and mouth opening of both participants and clinical interviewers were extracted from the first 3 minutes of a face-to-face, video-recorded Columbia-Suicide Severity Rating Scale (C-SSRS) using the Python-Based Automated Facial Affect Recognition (PyAFAR) software. Nonverbal behaviors of suicidal versus nonsuicidal participants and their interviewers were compared using 2-tailed independent samples t tests and Mann-Whitney U tests. Binary classification algorithms were then used to test how well these nonverbal behaviors together predicted group membership. Exploratory post hoc analyses assessed whether any nonverbal behaviors at baseline were associated with suicidal participants’ ideation severity or suicidal behavior 3 months later. Results: Nonverbal behaviors of participants and particularly their clinical interviewers differentiated suicidal versus nonsuicidal young adults at baseline. Suicidal participants demonstrated elevated velocity in opening and closing of eyes and mouth ( P =.004). Interviewers of suicidal participants showed less animated head movement ( P =.02), elevated velocity of eye opening and closing ( P =.02), and ambivalent smiling patterns ( P s=.01-.046). Overall, interviewer nonverbal behaviors predicted group membership with greater accuracy than participant behaviors, correctly identifying 81% (27/33) versus 59% (19/33) of suicidal young adults. Finally, interviewer smiling occurrence was associated with suicidal participants’ ideation severity ( P =.04) and suicidal behavior ( P =.02) 3 months later, while explicit measures, including interviewers’ clinical ratings and participants’ own self-reported ideation severity at baseline, were not. Conclusions: This study demonstrates the importance of attending to nonverbal channels of communication in suicide assessments, especially those of the clinical interviewer. It also highlights the potential for automated coding to detect clinically meaningful information efficiently and objectively.

BACKGROUND The importance of mental health has been increasingly highlighted, yet many individuals still face barriers to accessing suitable interventions. Although AI-based dialogue systems for mental health enhancement have advanced notably to address this issue, comprehensive surveys in this area, particularly those considering studies that adopt large language models (LLMs), remain scarce. OBJECTIVE This study aims to conduct a quantitative and qualitative review of current research trends in AI-driven dialogue systems for enhancing mental health. METHODS This study performed a bibliometric analysis and a trend review analysis of AI-driven dialogue systems for mental health, covering literature from 2020 to May 2024 across three citation databases—WoS, Scopus, and ACM Digital Library. The bibliometric analysis statistically assessed the distribution of publications, while the qualitative trend review focused on three key areas: (i) highly cited publications, (ii) those using the ESConv dataset, and (iii) those employing LLMs. RESULTS We reviewed 146 papers published between 2020 and 2024, observing a steady increase in publications over the last five years. Our bibliometric analysis examined publication distribution across sources, countries, institutions, and authors, while keyword network analysis highlighted major themes. Most of the top 10 highly cited papers focused on empathetic response generation, incorporating psychological approaches within deep learning models. For the ESConv dataset’s application in counseling, prominent techniques included multi-task learning and the integration of external knowledge. Lastly, we identified notable advantages of LLMs over traditional deep learning models and explored strategies to overcome their limitations as counseling tools. CONCLUSIONS Our study identifies key areas for developing counseling dialogue systems, such as incorporating psychological knowledge, improving data access, applying LLMs, and refining evaluation methods. By examining current research trends and establishing a foundational framework, this work offers future directions to enhance the effectiveness of AI counseling systems, contributing to both the machine learning and psychology fields.

Bipolar disorder (BD) is closely associated with an increased risk of suicide. However, while the prior work has revealed valuable insight into understanding the behavior of BD patients on social media, little attention has been paid to developing a model that can predict the future suicidality of a BD patient. Therefore, this study proposes a multi-task learning model for predicting the future suicidality of BD patients by jointly learning current symptoms. We build a novel BD dataset clinically validated by psychiatrists, including 14 years of posts on bipolar-related subreddits written by 818 BD patients, along with the annotations of future suicidality and BD symptoms. We also suggest a temporal symptom-aware attention mechanism to determine which symptoms are the most influential for predicting future suicidality over time through a sequence of BD posts. Our experiments demonstrate that the proposed model outperforms the state-of-the-art models in both BD symptom identification and future suicidality prediction tasks. In addition, the proposed temporal symptom-aware attention provides interpretable attention weights, helping clinicians to apprehend BD patients more comprehensively and to provide timely intervention by tracking mental state progression.

BACKGROUND Although mental illness is a leading cause of death, social stigma makes people hesitant to receive timely medical treatment. Online diagnosis systems can increase the accessibility of mental health care services. However, explaining illness detection results has been challenging in these systems because existing online diagnosis systems rely mainly on users’ self-assessments based on structured questionnaires. OBJECTIVE We proposed a system that provides detailed explanations (i.e., symptom evidence and solutions) about detected mental illnesses by analyzing users’ free-text inputs. METHODS Our system consists of three modules: a detection module, symptom extraction module, and solution suggestion module. The detection module analyzes the input text that the user describes, and accurately classifies the mental disorders that the user is likely to have. The system then extracts and delivers symptom evidence and solutions to explain the detection results better. A within-subject study was conducted to examine the effectiveness of the proposed system. Qualitative and quantitative data were collected using 36 questions, including seven open-ended questions, from 30 participants. RESULTS Analysis of the quantitative results demonstrated that additional symptom and solution information, along with the classification results, helped increase the explainability and usability of the online mental illness detection system. A total of 83.33% of the users preferred the system, which provided various explanations, and the free-text entry method helped them think more about themselves than answering structured questions. CONCLUSIONS The implementation of our proposed system revealed that (i) the system achieved fairly accurate performance in mental illness detection, (ii) providing explanations could improve user experience, and (iii) most users were satisfied with the system, while a few showed some concerns.

In general, there are several types of feedback provided in the game, and game users feel the fun of the game and use it continuously through such feedback. In this study, in order to minimize the defection of game users, a study was conducted to find out what customized feedback should be provided to users. To find out how motivation and positive emotions was expressed according to the level of achievement of users and the type of feedback, the goal pursuit methods were applied to the game as feedback, and the actual game was produced and analyzed by dividing the group according to the level of achievement. As a result, high scorers—professionals—have 'committed' feedback that gives them an overall score so they can check their efforts, and 'progress' feedback that lets them know they're reaching their goals, with similar levels of motivation and emotion in all conditions. Although they showed a positive emotional level, the low scorers, who can be called novices, showed a low level of motivation in the commitment feedback and a high level of motivation in the progress feedback. As a result of this study, it was discussed that different types of feedback should be presented according to the types of users.

Journal of Medical Internet Research - International Scientific Journal for Medical Research, Information and Communication on the Internet #Preprint #PeerReviewMe: Warning: This is a unreviewed preprint. Readers are warned that the document has not been peer-reviewed by expert/patient reviewers or an academic editor, may contain misleading claims, and is likely to undergo changes before final publication, if accepted, or may have been rejected/withdrawn. Readers with interest and expertise are encouraged to sign up as peer-reviewer, if the paper is within an open peer-review period. Please cite this preprint only for review purposes or for grant applications and CVs (if you are the author). No abstract/not applicable

UNSTRUCTURED No abstract/not applicable