Frontiers in Psychiatry

Mental health research increasingly focuses on the relationship between psychiatric symptoms and observable manifestations of the face and body1. In recent studies2,3, psychiatric patients have shown distinct patterns in movement, posture and facial expressions, suggesting these elements could enhance clinical diagnostics. The analysis of the facial expressions is grounded on the Facial Action Coding System (FACS)4. FACS provides a systematic method for categorizing facial expressions based on specific muscle movements, enabling detailed analysis of emotional and communicative behaviors. This method combined with recent advancements in Artificial Intelligence (AI) has shown promising results for the detection of the patient mental state. We analyze video data from patients with various psychiatric symptoms, using open-source Python toolboxes for facial expression and body movement analysis. These toolboxes facilitate face detection, facial landmark detection, emotion detection and motion recognition. Specifically, we aim to explore the connection between these physical expressions and established diagnostic tools, like symptom severity scores, and finally enhance psychiatric diagnostics by integrating AI-driven analysis of video data. By providing a more objective and detailed understanding of psychiatric symptoms, this study could lead to earlier detection and more personalized treatment approaches, ultimately improving patient outcomes. The findings will contribute to the development of innovative diagnostic tools that are both efficient and accurate, addressing a critical need in mental health care.

BackgroundSubstance use disorders (SUDs) continue to be a public health challenge of significant importance. The immunomodulatory effects of substances of abuse have been extensively studied but there is a dearth of information on their effects on plasma interleukin-40 (IL-40) level, a biomarker of B cell activity, and its consequent effects on plasma total IgA level in patients with substance use disorders (SUD). Therefore, the plasma levels of IL-40 and total IgA in SUD patients were determined in this study. MethodsNinety adults comprising 50 SUD patients and 40 controls were enrolled into this case-control study. The SUD patients were stratified into groups based on the number of substances they abuse and plasma levels of IL-40 and IgA were determined using ELISA. ResultsMarijuana was the most abused substance (68.0%) and majority of the SUD patients (64.0%) were polydrug users. The median plasma IL-40 level was significantly lower in SUD patients compared with the controls. Similarly, the median plasma total IgA level was significantly lower in SUD patients compared with the controls. However, there were no significant differences in the plasma levels of IL-40 and IgA in SUD patients who abuse single substance, two substances, and three or more substances. The plasma IL-40 level had significant positive correlation with IgA in SUD patients. ConclusionSubstance use disorder is associated with impaired humoral immune function, but the dysregulation appears not to be influenced by poly-drug use. Studies evaluating the mechanisms underlying humoral immune impairment in patients with substance use disorder and its potential clinical implications are suggested.

BackgroundHeroin and other opioid use disorders (HUD and OUD) cause massive public health morbidity and mortality. Although standard-of-care medication assisted treatment (MAT) exists, little is known about potential predictors of change during recovery. Recovery can include normalization of the brains white matter (WM) microstructure, which is sensitive to cytokine and immune signaling. Here we aimed to determine whether blood-based cytokine/immune markers can predict WM microstructure recovery following medication-assisted treatment. MethodsInpatient Individuals with HUD (iHUD; n=21) and healthy controls (HC; n=24) underwent magnetic resonance scans with diffusion tensor imaging (DTI) and provided ratings of drug cue-induced craving, arousal and valence twice, earlier in treatment and {approx}14 weeks of inpatient MAT (with methadone or buprenorphine) thereafter. At this second session (MRI2), they also provided a peripheral blood sample for multiplex relative quantification of serum cytokine/immune proteins (with a proximity extension assay, Olink). We explored the correlation of a multi-target cytokine biomarker score (based on principal component analysis of 19 proteins that differed significantly between iHUD and HC) with change in whole-brain DTI ({Delta}DTI; MRI2 - MRI1) metrics (fractional anisotropy, mean diffusivity, and axial diffusivity) across the 14 weeks of MAT. ResultsThe cytokine biomarker score, obtained at the MRI2 stage, was correlated with {Delta}DTI metrics in frontal, fronto-parietal, and cortico-limbic WM tracts (e.g., including the genu of the corpus callosum, anterior corona radiata, and others). In a follow-up analysis, specific cytokines represented in the multi-target biomarker score, such as the interleukin oncostatin M (OSM), colony stimulating factor (CSF21), and the chemokine CCL7 were correlated with similar {Delta}DTI metrics in iHUD, but not in HC. Levels of other specific cytokines (i.e., CCL19 and CCL2) were negatively correlated with change in cue-induced craving or arousal. Thus, lower levels of the aforementioned cytokines were correlated with an increase in cue-induced craving or arousal across the two stages (MRI2 - MRI1). ConclusionsStudied as a multi-target biomarker score, or as individual targets, peripheral serum cytokines are highly accessible biomarkers of WM microstructure recovery in iHUD undergoing inpatient MAT.

BackgroundThis study addresses limitations of traditional medication adherence assessment tools by developing a machine learning model to evaluate post-discharge medication compliance in patients using drugs. The research was conducted at Nanjing Drum Tower Hospital from February 2024 to December 2024. MethodsWe collected clinical data from 240 patients through questionnaires and developed a multi-class machine learning model. Feature selection employed manual screening and polynomial logistic regression. Six ML models were evaluated, with the Random Forest Classifier (RFC) demonstrating optimal performance (bad_AUC = 0.979, fine_AUC = 0.973, good_AUC = 0.917). SHAP analysis was used to explain the best-performing model. ResultsThe RFC model showed superior predictive capability across all adherence levels. Model interpretation revealed key clinical factors influencing adherence patterns. The tool enables early identification of non-compliance and supports intervention strategies. ConclusionsThis RFC-based model represents a significant advancement in medication adherence assessment, offering clinicians a practical tool for monitoring compliance. The approach shows particular promise for enhancing mental health management in this patient population by fostering better medication awareness and establishing scientific medication habits during early treatment stages.

This study aimed to assess job strain" (JS) and effort-reward imbalance (ERI), two markers of work-related stress, among psychiatric healthcare workers exposed to patients psychological suffering. A self-administered questionnaire was distributed in a regional French public hospital. The response rate was 65% and the prevalence of JS was estimated at 37%. Occurrence of JS was increased in the inpatient sector (OR 1.94 [1.01-3.74], P({chi}{superscript 2})=0.046), decreased by day work (OR 0.53 [0.29-0.95], P({chi}{superscript 2})=0.032), and by larger hospital seniority. The ERI ratio (mean 0.537 {+/-} 0.228) was higher among psychiatrists and workers doing overtime. It was lower in staff who felt supported during episodes of violence. These results suggest the need for specific preventive measures targeting these risk factors.

Background and ObjectiveAdvancement in mental health care requires easily accessible diagnostic and treatment assessment tools. There is an ongoing search for biomarkers that would enable objectification and automatization of the diagnostic and treatment process dependent on a psychiatric interview. Current wearable technology and computational methods make it possible to incorporate heart rate variability (HRV), an indicator of autonomic nervous system functioning and a potential biomarker of disease severity in mental disorders, into accessible diagnostic and treatment assessment frameworks. MethodsWe used a commercially available electrocardiography (ECG) chest strap with a built-in accelerometer, i.e. Polar H10, to record R-R intervals and activity of 30 hospitalized schizophrenia or bipolar disorder patients and 30 control participants for 1.5-2 hours time periods. We performed an analysis to assess the relationship between HRV and the Positive and Negative Syndrome Scale (PANSS) test scores. The source code for the reproduction of all experiments is available on GitHub while the dataset is available in Zenodo. Results and ConclusionsMean HRV values were lower in the patient group and negatively correlated with the results of the PANSS general subcategory. For the control group, we also discovered the inversely proportional dependency between the mobility coefficient based on accelerometer data and HRV. This relationship was less pronounced for the patient group. This indicates that HRV and mobility may be promising markers in disease diagnosis.

Borderline personality disorder (BPD) is characterized by an instability of self-image, interpersonal relationships, and emotions and is highly comorbid with other disorders. Physical activity has shown great results in treating these disorders. Physical activity intervention should be built considering the preferences and barriers of the targeted individuals. However, to this day no study analyzed the preferences and barriers to physical activity in individuals with BPD, which is the goal of this study. We used an online survey to question 192 adults with a self-reported diagnosis of BPD from Canada, France, the United States, England, Switzerland, and New Zealand. Participants complete 607 minutes of physical activity weekly on average. They prefer walking (66.7%), biking (33.3%), aquatic activities (29.0%), and running (24.2%). Their main barriers to physical activity are having a friend over, having other engagements, and recovering from an injury. They also prefer doing individual supervised physical activity outside and in a long session of moderate intensity. Finally, a majority of participants are interested in receiving physical activity advice, but most did not. The professionals from whom they would prefer to receive advice are trainers, psychiatrists, physical therapists, and psychologists. These results are important to better tailor future physical activity interventions for adults with BPD.

BACKGROUNDCognitive bias modification (CBM) programs have shown promise in treating psychiatric conditions, but they can be perceived as boring and repetitive. Incorporating gamified designs and adaptive algorithms in CBM training may address this issue and enhance engagement and effectiveness. OBJECTIVETo gather preliminary data and assess the preliminary efficacy of an adaptive approach bias modification (A-ApBM) paradigm in reducing cue-induced craving in individuals with methamphetamine use history. METHODSA randomized controlled trial with three arms was conducted. Individuals aged 18-60 with methamphetamine dependence and at least one year of methamphetamine use were recruited from 12 community-based rehabilitation centers in Sichuan, China. Individuals with inability to fluently operate a smartphone and/or the presence of mental health conditions other than methamphetamine use disorder (MUD) were excluded. A-ApBM group engaged in ApBM training using a smartphone application for four weeks. A-ApBM used an adaptive algorithm to dynamically adjust the difficulty level based on individual performance. Cue-induced craving scores and relapse were assessed using a visual analog scale at baseline, post-intervention, and at week-16 follow-up. RESULTSA total of 136 participants were recruited and randomized: 48 were randomized to the A-ApBM group, 48 were randomized to the S-ApBM group, and 40 were randomized to the no-intervention control group. The A-ApBM group showed a significant reduction in cue-induced craving scores at post-intervention compared to baseline (Cohens d = 0.34, p < 0.01, 95% CI = [0.03,0.54]). The reduction remained significant at the week-16 follow-up (Cohens d = 0.40, p= 0.01, 95% CI = [0.18,0.57]). No significant changes were observed in the S-ApBM and control groups. CONCLUSIONThe adaptive ApBM paradigm with gamified designs and dynamic difficulty adjustments may be an effective intervention for reducing cue-induced craving in individuals with methamphetamine use history. This approach improves engagement and personalization, potentially enhancing the effectiveness of CBM programs. Further research is needed to validate these findings and explore the application of adaptive ApBM in other psychiatric conditions. TRIAL REGISTRATIONRegistered at clinicaltrials.gov (ID NCT05794438).

Mood instability and circadian rhythm disruptions are both of increasing interest with regard to a number of psychiatric disorders, notably bipolar disorder (BD), but understanding of their nature and their interrelationship are incomplete. By definition, both have an integral temporal component and, as such, measuring them longitudinally and remotely is desirable. We conducted the Cognition and Mood Evolution across Time (COMET) study to assess the feasibility and value of digital devices to capture mood, its instability, and daily rest-activity patterns, over a 10-week period, in two groups of participants. The first group (n=37) were selected as scoring >7 on the Mood Disorder Questionnaire (MDQ) ( high MDQ), thereby having a history of mood elevation and being at risk for BD. They were compared with a group (n=37) scoring <5 on the MDQ ( low MDQ). Over a 10-week period, using a tablet, mood was rated daily, clinical ratings of depression, mania, and anxiety were captured weekly via the True Colours app, and a GENEactiv actigraph was worn to capture rest-activity pattern data. The main findings are that (1) MDQ score predicts mood instability; (2) high MDQ score is associated with more negative affect and mood symptoms than people with low MDQ, and with a different circadian activity profile; and (3) mood instability and circadian indices appear uncorrelated. The implications are that (1) remote monitoring of these domains is feasible and valuable; (2) selection of participants based on MDQ score is useful for studying mood (in)stability; and (3) the approach has potential for studies of clinical populations and for experimental medicine studies assessing interventions to reduce mood instability.

BackgroundUniversal access to assessment and treatment of mental health and learning disorders remains a significant and unmet need. There is a vast number of people without access to care because of economic, geographic, and cultural barriers as well as limited availability of clinical experts who could help advance our understanding of mental health. ObjectiveTo create an open, configurable software platform to build clinical measures, mobile assessments, tasks, and interventions without programming expertise. Specifically, our primary requirements include: an administrator interface for creating and scheduling recurring and customized questionnaires where end users receive and respond to scheduled notifications via an iOS or Android app on a mobile device. Such a platform would help relieve overwhelmed health systems, and empower remote and disadvantaged subgroups in need of accurate and effective information, assessment, and care. This platform has potential to advance scientific research by supporting the collection of data with instruments tailored to specific scientific questions from large, distributed, and diverse populations. MethodsWe conducted a search for tools that satisfy the above requirements. We designed and developed a new software platform called "MindLogger" that exceeds the above requirements. To demonstrate the tools configurability, we built multiple "applets" (collections of activities) within the MindLogger mobile application and deployed several, including a comprehensive set of assessments underway in a large-scale, longitudinal, mental health study. ResultsOf the hundreds of products we researched, we found 10 that met our primary requirements above with 4 that support end-to-end encryption, 2 that enable restricted access to individual users data, 1 that provides open source software, and none that satisfy all three. We compared features related to information presentation and data capture capabilities, privacy and security, and access to the product, code, and data. We successfully built MindLogger mobile and web applications, as well as web browser-based tools for building and editing new applets and for administering them to end users. MindLogger has end-to-end encryption, enables restricted access, is open source, and supports a variety of data collection features. One applet is currently collecting data from children and adolescents in our mental health study, and other applets are in different stages of testing and deployment for use in clinical and research settings. ConclusionsWe have demonstrated the flexibility and applicability of the MindLogger platform through its deployment in a large-scale, longitudinal, mobile mental health study, and by building a variety of other mental health-related applets. With this release, we encourage a broad range of users to apply the MindLogger platform to create and test applets to advance health care and scientific research. We hope that increasing availability of applets designed to assess and administer interventions will facilitate access to health care in the general population.

Substance use disorder (SUD) recovery typically requires transformative change and prioritizing long-term healthy goals. Unfortunately, successful recovery is threatened by relapse rates that often exceed 50% in the first year. We previously reported on an experiential virtual reality (VR) SUD recovery intervention using personalized future self-avatars that produced emotional engagement and positive behavioral change, ie, stronger connection with the future self and future rewards and reduced craving. Here, we used fMRI to identify brain engagement to a future self experience with divergent futures. Twenty adults (14 male, 33 years old) in early SUD recovery (<1 year) interacted with age-progressed versions of themselves in two different VR future realities: an SUD Future Self and a Recovery Future Self. Vivid lifelike visual and audio animation was augmented with a personalized narrative concerning future drug use and recovery. MRI immediately followed. Participants viewed videos of their future selves in the virtual environment and were directed to contemplate what they were seeing. Viewing and contemplating the future selves elicited activation in midline default mode regions (posterior cingulate and ventromedial prefrontal cortices), visual regions including the occipital and fusiform face areas, and left middle frontal gyrus. The Recovery Future Self produced significant left occipital face area activation compared with the SUD Future Self. Midline default mode activation correlated with VR-induced increases in delayed reward preference, and also with greater trait perseverance. Using digital selves as therapeutic agents reveals an entirely novel set of possible interventions and opens exciting new frontiers in behavior change methodology. Future studies targeting decision-making and future behavior could be informed by evaluating increased midline default mode engagement, with uniquely self-focused mechanisms signaled by executive network and face area coactivation. New hope for treatment-resistant mental health conditions is offered by the nearly limitless range of therapeutic experiences enabled by immersive digital therapeutics. Plain Language SummaryHigh relapse rates in early recovery remains a serious challenge. To promote better outcomes, our team recently developed a virtual reality experience where people interacted with future versions of themselves. We used magnetic resonance imaging (MRI) to understand how the brain activated to this experience, and what brain responses were linked to positive outcomes. We worked with 20 adults in early recovery. Each person used virtual reality to interact with two different future selves: one who had returned to substance use, and one who had stayed in recovery. These digital future selves looked and sounded like the participants and were paired with a personalized story about future drug use and recovery. Right after the virtual reality session, participants brains were scanned while they watched videos of these future selves and were asked to think about what they were seeing. When people viewed and reflected on their future selves, brain areas involved in self-reflection and imagining the future became more active, along with regions that process faces. The future selves triggered brain activation in "self-focused" brain networks and in face-processing regions. Activity in key "self-focused" brain regions was linked to choosing larger, delayed rewards over smaller, immediate ones, and to lower impulsivity. These findings suggest that lifelike digital versions of peoples future selves engage brain systems that support thinking ahead, persistence, and valuing long-term outcomes. This creates a promising new avenue for immersive digital therapeutic experiences to encourage lasting behavior change in early recovery from substance use disorder.

IntroductionCOVID-19 outbreak has imposed an eight-week confinement in France. During this period, children and their families were exposed to a full-time home life. The aim of this study was to assess the emotional experience and tolerance of children with autism spectrum disorder (ASD) in this particular context. MethodA clinical survey was proposed to parents and rated by professionals once a week during the quarantine period in France. 95 autistic children followed by the child and adolescent psychiatry department of Tours university hospital were assessed from the 18th of March to the 8th of May. The following clinical points were investigated: child anxiety, family anxiety, behavior problems, impact on sleep, impact on appetite, impact on school work, family tension, confinement intolerance, difficulties to follow a schedule, isolation behavior. ResultsDespite minor changes in family anxiety and school work, no difference was highlighted between clinical scores collected at the beginning and at the end of this period. ASD children with or without intellectual disability had non-significant clinical changes during quarantine. This evolution was also independent of the accommodation type (house or apartment) and the parental status (relationship, separated or isolated). ConclusionThe sameness dimension in autism and parents adaptation may be involved in this clinical stability during COVID confinement. Moreover, specialized tools and support provided by professionals could have participated to these outcomes and must be regularly promoted in order to help families in this still difficult period.

PurposeAccurate psychiatric diagnosis and assessment are crucial for effective treatment. However, while current data-driven approaches emphasize diagnostic outcomes, the process of decoding the underlying symptom expressions in patients language and mapping them to well-defined psychiatric terminology has received relatively little attention. This study investigates the potential of Large Language Models (LLMs) to automate the identification of diagnostic categories and symptoms from psychiatrist-patient dialogues, to provide interpretable insights and support automatic diagnosis. MethodsWe analyzed audio recordings from 1160 psychiatric diagnostic interviews, primarily involving patients with depressive disorder and anxiety disorder. A clinical entities corpus was formed by leveraging clinical annotations in EMRs (e.g., chief complaints, mental status, elements in assessment scales) and widely used assessment scales. LLMs were utilized to identify clinical symptoms, rate assessment scales, and an ensemble learning pipeline was designed to classify diagnostic results and symptoms with 10-fold cross-validation. ResultsThe system achieved 86.9% accuracy for identifying the appearance of clinical annotations and 74.7% (77.2%) accuracy for identifying anxiety (depression) symptoms. Patients with depression and anxiety, diagnosed using ICD-10 codes, were differentiated with an accuracy of 75.5%. Analysis of LLM-generated features shows that depression cases exhibited prominent markers of anhedonia and decreased volition, whereas anxiety disorders were characterized by tension and an inability to relax. ConclusionThis study demonstrates the potential of integrating LLM technology with linguistic and acoustic features to enhance psychiatric diagnostics. The developed pipeline effectively predicts psychiatric diagnoses and provides interpretable insights, showcasing a valuable tool for clinicians in mental health assessment.

Mental health is in a state of crisis with demand for mental health services significantly surpassing available care. As such, building scalable and objective measurement tools for mental health evaluation is of primary concern. Given the usage of spoken language in diagnostics and treatment, it stands out as potential methodology. Here a model is built for mental health status evaluation using natural language processing. Specifically, a RoBERTa-based model is fine-tuned on text from psychotherapy sessions to predict mental health status with prediction accuracy on par with clinical evaluations at 74%.

Irritability and temper outbursts are among the most common reasons youth are referred for psychiatric assessment and care. Identifying clinical variables (e.g. momentary anxiety) that precede the onset of temper outbursts would provide valuable clinical utility. Here, we provide the rationale for a study to test the performance of classifiers trained to predict temper outbursts in a group of clinically-referred youth, in a home setting, enriched for symptoms of irritability and temper outbursts. Using observational data--digital based event sampling from previous Ecological Momentary Assessment data, we demonstrate promising results in our ability to predict the presence of a temper outburst based on clinical responses (e.g., whether the participant is grouchy, hungry, happy, sad, anxious, tired, etc.) prior to the emotional event, as well as external features (e.g., time of day, day of week). In exploratory analyses of existing data, consisting of n=57 subjects with a total of 1296 time points, we evaluate the feasibility of using a logistic regression-based classifier and a random-forest based classifier for predicting the temper outburst prospectively. In order to more rigorously assess these classifiers, we propose the collection of a large confirmatory set, consisting of at least an additional 20 subjects with an expected total of 400 time points, and will perform confirmatory analyses of the precision and recall of several classifiers for predicting temper outbursts. This work provides the foundation for the identification of features predictive of risk and future development of novel mobile-device-based interventions in youth affected with severe and impairing psychopathology.

Internet gaming disorder (IGD) presents a significant public health challenge, yet its complex biopsychosocial mechanisms and dynamic risk trajectories remain poorly understood due to a scarcity of comprehensive longitudinal and multimodal cohorts. To address this critical gap, we established the Chinese College Student Gamers Cohort (CCSGC), a prospective, multimodal longitudinal study of 793 first-year undergraduates primarily playing Honor of Kings from 2022 Sept. The CCSGC integrates semi-annual psychosocial questionnaires, annual neuroimaging (EEG/fMRI), and biospecimen collection over multiple years. Baseline data revealed individuals with IGD (n=211) exhibited significantly higher gaming craving, psychological distress (depression, anxiety), impulsivity, and maladaptive motivational features compared to non-IGD gamers (regular players (RP) n=400; casual players (CP) n=182). Longitudinal analyses across four waves indicated bidirectional temporal associations between IGD severity and mental symptoms, and a stabilization of IGD incidence after an initial decrease. Furthermore, specific neurophysiological (e.g., N400 amplitude to game cues) and neuroimaging (e.g., superior parietal activation) markers were identified that correlated with IGD severity and predicted one-year outcomes in gaming disorder or social functioning. The CCSGC provides an invaluable resource for dissecting the heterogeneity, comorbidity, and intricate biopsychosocial mechanisms of IGD, holding significant potential to advance risk prediction, early identification, and targeted intervention strategies.

The COVID-19 pandemic has presented a significant challenge to societal mental health. Yet, it remains unknown which factors influence the mental adaptation from lockdown to subsequent relaxation periods, particularly for vulnerable groups. This study used smartphone-based monitoring to explore how 74 individuals with major depression (MDD) and 77 healthy controls (HCs) responded to the transition from lockdown to relaxation during the first wave of the COVID-19 pandemic (March 21 to November 01, 2020) regarding interpersonal interactions, COVID-19-related fear (fear of participants own health, the health of close relatives, and the pandemics economic impact), and the feeling of isolation. Furthermore, we investigated the effect of a diagnosis of MDD and the experience of childhood maltreatment (CM) on adaptive functioning. During the transition from lockdown to relaxation, we observed an increase in direct contacts and a decrease in indirect contacts and self-perceived isolation in the study population. The diagnosis of MDD and the experience of CM moderated a maintenance of COVID-19-related fear: HCs and participants without the experience of CM showed a decrease in fear, while fear of participants with MDD and with an experience of CM did not change significantly. The finding that elevated COVID-19-related fear was sustained in vulnerable groups after lockdown measures were lifted could help guide psychosocial prevention efforts in future pandemic emergencies.

BackgroundCocaine Use Disorder (CUD) poses significant neurobiological and neuropsychiatric challenges, often resulting in severe cognitive and behavioral impairments. This study aims to explore the neural dynamics of CUD using a dynamic coactivation pattern (CAP) analysis approach to provide a deeper understanding of the transient neurobiological mechanisms of the disorder. MethodsResting-state functional MRI data (SUDMEX_CONN) from 56 CUD patients and 57 healthy controls (HC) were analyzed. CAP analysis was employed to capture transient brain states and their coactivation patterns. Temporal dynamic metrics such as Fraction of Time, Persistence (PST), and Counts were computed to assess differences between groups. Stationary functional connectivity (sFC) was also examined, and meta-analytic term mapping from the Neurosynth database was used to characterize functional associations. ResultsCAP analysis revealed six distinct coactivation patterns, with five showing high spatial similarity between CUD and HC groups. Notable differences were observed in State 6, which displayed inverse activation patterns between the groups. CUD individuals exhibited significantly reduced PST across all brain states and altered transition probabilities, particularly increased transitions from the default mode network (DMN) to the somatomotor network and decreased transitions from DMN to attentional/executive networks. Clinical correlations indicated that prolonged cocaine use was associated with altered PST in specific brain states. sFC analysis identified significant alterations in regions such as the right supramarginal gyrus, left superior frontal gyrus, right precentral gyrus, and right lingual gyrus, each linked to distinct cognitive and behavioral functions. ConclusionsThis study highlights the utility of CAP analysis in capturing the dynamic neural underpinnings of CUD. The findings provide insights into the neurobiological mechanisms of the disorder, suggesting potential biomarkers for CUD. These results have implications for developing an enhanced approach for substance use disorders, as well as improving our understanding and management of CUD.

BackgroundMental disorders affect approximately 14% of adolescents worldwide, often leading to persistent cognitive and emotional difficulties and reduced health-related quality of life (HRQoL). Executive functions (EF)--including cognitive flexibility, inhibition, attention, and working memory-- are particularly impaired in many psychiatric conditions. Chess has recently been proposed as a low-cost cognitive remediation training (CRT). This pilot study investigated whether chess-based CRT could enhance EF and HRQoL in adolescents with psychiatric disorders. MethodsA quasi-experimental design was conducted at the Department of Child and Adolescent Psychiatry, Mannheim, Germany (September 2022-April 2024). Participants aged 13-17 years were assigned to either a six-week chess intervention (experimental group, EG) or treatment as usual (control group, CG). Both groups received standard multidisciplinary therapy, while the EG additionally participated in weekly 90-minute chess sessions based on The Kings Plan for Kids. Cognitive flexibility (DCCS), inhibitory control (Stop-Signal Task), sustained attention (d2-R), and working memory (n-back task) were assessed alongside HRQoL (KIDSCREEN-27). Data were analyzed using t-tests. ResultsThirty-three adolescents were included (19 EG, 14 CG; 82% female). While no significant group differences emerged for cognitive flexibility, inhibitory control, or sustained attention, the EG showed significantly faster reaction times in the working memory task (p = .016, d = 0.79), suggesting improved cognitive efficiency. Psychological well-being increased significantly in the EG compared to the CG (p = .035, d = 0.67), whereas physical well-being showed a non-significant upward trend. ConclusionChess-based CRT was associated with enhanced psychological well-being and improved working memory efficiency in adolescents with psychiatric disorders. Although other EF measures did not show significant changes, findings support the feasibility and potential clinical value of chess as an engaging, low-risk adjunct to standard therapy. Larger randomized trials are needed to confirm these preliminary results.

Artificial intelligence (AI) is increasingly leveraged in mental healthcare for early detection, monitoring, and personalized intervention. However, most existing AI applications are based on categorical diagnostic systems like DSM-5 or ICD-11, which often lead to comorbidity issues, ambiguous diagnoses, and insufficient personalization. These tools typically target specific disorders (e.g., depression or anxiety), neglecting the broader, interconnected nature of psychopathological symptoms. Addressing these limitations, recent innovations in psychopathology emphasize transdiagnostic and network-based approaches, such as the Hierarchical Taxonomy of Psychopathology (HiTOP), which conceptualize mental disorders as dimensional and inter-connected constructs. This study proposes an AI-powered tool that integrates data-driven principles from both the HiTOP and symptom network models to generate individualized risk profiles for internalizing mental disorders (e.g., depression, anxiety, bipolar disorders). Our solution aims to assess individuals current psychopathological traits and symptom components, providing a comprehensive, nuanced profile supporting clinical diagnoses and monitoring. The study unfolds in three phases: (1) model ideation; (2) model implementation in a large-scale Vietnamese sample; and (3) deployment in clinical and psychological practice settings in Vietnam. Central to our method is the development of a Risk-aware Taxonomy-enhanced Symptom Encoder (RiTaSE), which encodes symptom data and their severities into rich representations processed via a Transformer-based model. The model is trained using high-quality, validated datasets mapped to the HiTOP framework. This project is among the first to employ AI for personalized psychopathological profiling in Vietnam,, as well as other low- and middle-income countries. Expected outcomes include an advanced diagnostic-support tool for clinical use, improved crosscultural insights into symptom comorbidity, and practical utility in mental health monitoring and intervention evaluation. Future extensions aim to broaden the scope across all HiTOP dimensions and predict transitions to clinical states through longitudinal and multi-modal data integration.