Identifying a clinical signature of suicidality among patients with mood disorders: A pilot study using a machine learning approach

Predicting inmate suicidal behavior with an interpretable ensemble machine learning approach in smart prisons

The convergence of smart technologies and predictive modelling in prisons presents an exciting opportunity to revolutionize the monitoring of inmate behaviour, allowing for the early detection of signs of distress and the effective mitigation of suicide risks. While machine learning algorithms have been extensively employed in predicting suicidal behaviour, a critical aspect that has often been overlooked is the interoperability of these models. Most of the work done on model interpretations for suicide predictions often limits itself to feature reduction and highlighting important contributing features only. To address this research gap, we used Anchor explanations for creating human-readable statements based on simple rules, which, to our knowledge, have never been used before for suicide prediction models. We also overcome the limitation of anchor explanations, which create weak rules on high-dimensionality datasets, by first reducing data features with the help of SHapley Additive exPlanations (SHAP). We further reduce data features through anchor interpretations for the final ensemble model of XGBoost and random forest. Our results indicate significant improvement when compared with state-of-the-art models, having an accuracy and precision of 98.6% and 98.9%, respectively. The F1-score for the best suicide ideation model appeared to be 96.7%.

Explainable artificial intelligence models for predicting risk of suicide using health administrative data in Quebec

Suicide is a complex, multidimensional event, and a significant challenge for prevention globally. Artificial intelligence (AI) and machine learning (ML) have emerged to harness large-scale datasets to enhance risk detection. In order to trust and act upon the predictions made with ML, more intuitive user interfaces must be validated. Thus, Interpretable AI is one of the crucial directions which could allow policy and decision makers to make reasonable and data-driven decisions that can ultimately lead to better mental health services planning and suicide prevention. This research aimed to develop sex-specific ML models for predicting the population risk of suicide and to interpret the models. Data were from the Quebec Integrated Chronic Disease Surveillance System (QICDSS), covering up to 98% of the population in the province of Quebec and containing data for over 20,000 suicides between 2002 and 2019. We employed a case-control study design. Individuals were considered cases if they were aged 15+ and had died from suicide between January 1st, 2002, and December 31st, 2019 (n = 18339). Controls were a random sample of 1% of the Quebec population aged 15+ of each year, who were alive on December 31st of each year, from 2002 to 2019 (n = 1,307,370). We included 103 features, including individual, programmatic, systemic, and community factors, measured up to five years prior to the suicide events. We trained and then validated the sex-specific predictive risk model using supervised ML algorithms, including Logistic Regression (LR), Random Forest (RF), Extreme Gradient Boosting (XGBoost) and Multilayer perceptron (MLP). We computed operating characteristics, including sensitivity, specificity, and Positive Predictive Value (PPV). We then generated receiver operating characteristic (ROC) curves to predict suicides and calibration measures. For interpretability, Shapley Additive Explanations (SHAP) was used with the global explanation to determine how much the input features contribute to the models' output and the largest absolute coefficients. The best sensitivity was 0.38 with logistic regression for males and 0.47 with MLP for females; the XGBoost Classifier with 0.25 for males and 0.19 for females had the best precision (PPV). This study demonstrated the useful potential of explainable AI models as tools for decision-making and population-level suicide prevention actions. The ML models included individual, programmatic, systemic, and community levels variables available routinely to decision makers and planners in a public managed care system. Caution shall be exercised in the interpretation of variables associated in a predictive model since they are not causal, and other designs are required to establish the value of individual treatments. The next steps are to produce an intuitive user interface for decision makers, planners and other stakeholders like clinicians or representatives of families and people with live experience of suicidal behaviors or death by suicide. For example, how variations in the quality of local area primary care programs for depression or substance use disorders or increased in regional mental health and addiction budgets would lower suicide rates.

Plasma and Platelet Brain-Derived Neurotrophic Factor (BDNF) Levels in Bipolar Disorder Patients with Post-Traumatic Stress Disorder (PTSD) or in a Major Depressive Episode Compared to Healthy Controls

Post-traumatic stress disorder (PTSD) is a highly disabling mental disorder arising after traumatism exposure, often revealing critical and complex courses when comorbidity with bipolar disorder (BD) occurs. To search for PTSD or depression biomarkers that would help clinicians define BD presentations, this study aimed at preliminarily evaluating circulating brain-derived-neurotrophic factor (BDNF) levels in BD subjects with PTSD or experiencing a major depressive episode versus controls. Two bloodstream BDNF components were specifically investigated, the storage (intraplatelet) and the released (plasma) ones, both as adaptogenic/repair signals during neuroendocrine stress response dynamics. Bipolar patients with PTSD (n = 20) or in a major depressive episode (n = 20) were rigorously recruited together with unrelated healthy controls (n = 24) and subsequently examined by psychiatric questionnaires and blood samplings. Platelet-poor plasma (PPP) and intraplatelet (PLT) BDNF were measured by ELISA assays. The results showed markedly higher intraplatelet vs. plasma BDNF, confirming platelets' role in neurotrophin transport/storage. No between-group PPP-BDNF difference was reported, whereas PLT-BDNF was significantly reduced in depressed BD patients. PLT-BDNF negatively correlated with mood scores but not with PTSD items like PPP-BDNF, which instead displayed opposite correlation trends with depression and manic severity. Present findings highlight PLT-BDNF as more reliable at detecting depression than PTSD in BD, encouraging further study into BDNF variability contextually with immune-inflammatory parameters in wider cohorts of differentially symptomatic bipolar patients.

Machine learning and the prediction of suicide in psychiatric populations: a systematic review

Machine learning (ML) has emerged as a promising tool to enhance suicidal prediction. However, as many large-sample studies mixed psychiatric and non-psychiatric populations, a formal psychiatric diagnosis emerged as a strong predictor of suicidal risk, overshadowing more subtle risk factors specific to distinct populations. To overcome this limitation, we conducted a systematic review of ML studies evaluating suicidal behaviors exclusively in psychiatric clinical populations. A systematic literature search was performed from inception through November 17, 2022 on PubMed, EMBASE, and Scopus following the PRISMA guidelines. Original research using ML techniques to assess the risk of suicide or predict suicide attempts in the psychiatric population were included. An assessment for bias risk was performed using the transparent reporting of a multivariable prediction model for individual prognosis or diagnosis (TRIPOD) guidelines. About 1032 studies were retrieved, and 81 satisfied the inclusion criteria and were included for qualitative synthesis. Clinical and demographic features were the most frequently employed and random forest, support vector machine, and convolutional neural network performed better in terms of accuracy than other algorithms when directly compared. Despite heterogeneity in procedures, most studies reported an accuracy of 70% or greater based on features such as previous attempts, severity of the disorder, and pharmacological treatments. Although the evidence reported is promising, ML algorithms for suicidal prediction still present limitations, including the lack of neurobiological and imaging data and the lack of external validation samples. Overcoming these issues may lead to the development of models to adopt in clinical practice. Further research is warranted to boost a field that holds the potential to critically impact suicide mortality.

The use of machine learning on administrative and survey data to predict suicidal thoughts and behaviors: a systematic review

Background

Machine learning is a promising tool in the area of suicide prevention due to its ability to combine the effects of multiple risk factors and complex interactions. The power of machine learning has led to an influx of studies on suicide prediction, as well as a few recent reviews. Our study distinguished between data sources and reported the most important predictors of suicide outcomes identified in the literature.

Objective

Our study aimed to identify studies that applied machine learning techniques to administrative and survey data, summarize performance metrics reported in those studies, and enumerate the important risk factors of suicidal thoughts and behaviors identified.

Methods

A systematic literature search of PubMed, Medline, Embase, PsycINFO, Web of Science, Cumulative Index to Nursing and Allied Health Literature (CINAHL), and Allied and Complementary Medicine Database (AMED) to identify all studies that have used machine learning to predict suicidal thoughts and behaviors using administrative and survey data was performed. The search was conducted for articles published between January 1, 2019 and May 11, 2022. In addition, all articles identified in three recently published systematic reviews (the last of which included studies up until January 1, 2019) were retained if they met our inclusion criteria. The predictive power of machine learning methods in predicting suicidal thoughts and behaviors was explored using box plots to summarize the distribution of the area under the receiver operating characteristic curve (AUC) values by machine learning method and suicide outcome (i.e., suicidal thoughts, suicide attempt, and death by suicide). Mean AUCs with 95% confidence intervals (CIs) were computed for each suicide outcome by study design, data source, total sample size, sample size of cases, and machine learning methods employed. The most important risk factors were listed.

Results

The search strategy identified 2,200 unique records, of which 104 articles met the inclusion criteria. Machine learning algorithms achieved good prediction of suicidal thoughts and behaviors (i.e., an AUC between 0.80 and 0.89); however, their predictive power appears to differ across suicide outcomes. The boosting algorithms achieved good prediction of suicidal thoughts, death by suicide, and all suicide outcomes combined, while neural network algorithms achieved good prediction of suicide attempts. The risk factors for suicidal thoughts and behaviors differed depending on the data source and the population under study.

Conclusion

The predictive utility of machine learning for suicidal thoughts and behaviors largely depends on the approach used. The findings of the current review should prove helpful in preparing future machine learning models using administrative and survey data.

Systematic review registration

https://www.crd.york.ac.uk/prospero/display_record.php?ID=CRD42022333454 identifier CRD42022333454.

Prediction of Repeated Self-Harm in Six Months: Comparison of Traditional Psychometrics With Random Forest Algorithm

Suicidal risk has been a significant mental health problem. However, the predictive ability for repeated self-harm (SH) has not improved over the past decades. This study thus aimed to explore a potential tool with theoretical accommodation and clinical application by employing traditional logistic regression (LR) and newly developed machine learning, random forest algorithm (RF). Starting with 89 items from six commonly used scales (i.e., proximal suicide risk factors) as preliminary predictors, both LR and RF resulted in a better solution with much fewer items in two phases of item selections and analyses, with prediction accuracy 88.6% and 79.8%, respectively. A combination with 12 selected items, named LR-12, well predicted repeated self-harm in 6-month follow-up with satisfactory performance (AUC = 0.84, 95% CI: 0.76-0.92; cut-off point by 1/2 with sensitivity 81.1% and specificity 74.0%). The psychometrically appealing LR-12 could be used as a screening scale for suicide risk assessment.

Blood epigenome-wide association studies of suicide attempt in adults with bipolar disorder

Suicide attempt (SA) risk is elevated in individuals with bipolar disorder (BD), and DNA methylation patterns may serve as possible biomarkers of SA. We conducted epigenome-wide association studies (EWAS) of blood DNA methylation associated with BD and SA. DNA methylation was measured at >700,000 positions in a discovery cohort of n = 84 adults with BD with a history of SA (BD/SA), n = 79 adults with BD without history of SA (BD/non-SA), and n = 76 non-psychiatric controls (CON). EWAS revealed six differentially methylated positions (DMPs) and seven differentially methylated regions (DMRs) between BD/SA and BD/non-SA, with multiple immune-related genes implicated. There were no epigenome-wide significant differences when BD/SA and BD/non-SA were each compared to CON, and patterns suggested that epigenetics differentiating BD/SA from BD/non-SA do not differentiate BD/non-SA from CON. Weighted gene co-methylation network analysis and trait enrichment analysis of the BD/SA vs. BD/non-SA contrast further corroborated immune system involvement, while gene ontology analysis implicated calcium signalling. In an independent replication cohort of n = 48 BD/SA and n = 47 BD/non-SA, fold changes at the discovery cohort's significant sites showed moderate correlation across cohorts and agreement on direction. In both cohorts, classification accuracy for SA history among individuals with BD was highest when methylation at the significant CpG sites as well as information from clinical interviews were combined, with an AUC of 88.8% (CI = 83.8-93.8%) and 82.1% (CI = 73.6-90.5%) for the combined epigenetic-clinical classifier in the discovery and replication cohorts, respectively. Our results provide novel insight to the role of immune system functioning in SA and BD and also suggest that integrating information from multiple levels of analysis holds promise to improve risk assessment for SA in adults with BD.

Prediction of suicidal ideation in children and adolescents using machine learning and deep learning algorithm: A case study in South Korea where suicide is the leading cause of death

BACKGROUND

Korea has the highest suicide rate among Organisation for Economic Co-operation and Development (OECD) countries. Consequently, central and local governments and private organizations in Korea cooperate in promoting various suicide prevention projects to actively respond to suicide problems. Machine learning has been used to predict suicidal ideation in the fields of health and medicine but not from a social science perspective.

OBJECTIVE

Since suicidal ideation is a major predictor of suicide attempts, being able to anticipate and mitigate it helps prevent suicide. Therefore, this study presents a data-based analysis method for predicting suicidal thoughts quickly and effectively and suggests countermeasures against the causes of suicidal thoughts.

PARTICIPANTS AND METHODS

To predict early signs of suicidal ideation in children and adolescents, big data collected for approximately 4 years (from 2017 to 2020) from the Korea Youth Policy Institute (NYPI) were used. To accurately predict suicidal ideation, supervised ma- chine learning classification algorithms such as logistic regression, random forest, XGBoost, multilayer perceptron (MLP), and convolutional neural network (CNN) were used.

RESULTS

Using CNN, suicidal ideation was predicted with an accuracy of approximately 90 %. The logistic regression results showed that sadness and depression increased suicidal thoughts by more than 25 times, and anxiety, loneliness, and experience of abusive language increased suicidal thoughts by more than three times.

CONCLUSIONS

Machine learning and deep learning approaches have the potential to predict and respond to suicidal thoughts in children, adolescents, and the general population, as well as help respond to the suicide crisis by preemptively identifying the cause.

Blood epigenome-wide association studies of suicide attempt in adults with bipolar disorder

Suicide attempt (SA) risk is elevated in individuals with bipolar disorder (BD), and DNA methylation patterns may serve as possible biomarkers of SA. We conducted epigenome-wide association studies (EWAS) of blood DNA methylation associated with BD and SA. DNA methylation was measured at > 700,000 positions in a discovery cohort of n = 84 adults with BD with a history of SA (BD/SA), n = 79 adults with BD without history of SA (BD/non-SA), and n = 76 non-psychiatric controls (CON). EWAS revealed six differentially methylated positions (DMPs) and seven differentially methylated regions (DMRs) between BD/SA and BD/non-SA, with multiple immune-related genes implicated. There were no epigenome-wide significant differences when BD/SA and BD/non-SA were each compared to CON, and patterns suggested that epigenetics differentiating BD/SA from BD/non-SA do not differentiate BD/non-SA from CON. Weighted gene co-methylation network analysis and trait enrichment analysis of the BD/SA vs. BD/non-SA contrast further corroborated immune system involvement, while gene ontology analysis implicated calcium signalling. In an independent replication cohort of n = 48 BD/SA and n = 47 BD/non-SA, fold-changes at the discovery cohort's significant sites showed moderate correlation across cohorts and agreement on direction. In both cohorts, classification accuracy for SA history among individuals with BD was highest when methylation at the significant CpG sites as well as information from clinical interviews were combined, with an AUC of 88.8% (CI = 83.8-93.8%) and 82.1% (CI = 73.6-90.5%) for the combined epigenetic-clinical predictor in the discovery and replication cohorts, respectively. Our results provide novel insight to the role of immune system functioning in SA and BD and also suggest that integrating information from multiple levels of analysis holds promise to improve risk assessment for SA in adults with BD.

An Overview of Bipolar Disorder Diagnosis Using Machine Learning Approaches: Clinical Opportunities and Challenges

Objective: Automatic diagnosis of psychiatric disorders such as bipolar disorder (BD) through machine learning techniques has attracted substantial attention from psychiatric and artificial intelligence communities. These approaches mostly rely on various biomarkers extracted from electroencephalogram (EEG) or magnetic resonance imaging (MRI)/functional MRI (fMRI) data. In this paper, we provide an updated overview of existing machine learning-based methods for bipolar disorder (BD) diagnosis using MRI and EEG data. Method : This study is a short non-systematic review with the aim of describing the current situation in automatic diagnosis of BD using machine learning methods. Therefore, an appropriate literature search was conducted via relevant keywords for original EEG/MRI studies on distinguishing BD from other conditions, particularly from healthy peers, in PubMed, Web of Science, and Google Scholar databases. Results: We reviewed 26 studies, including 10 EEG studies and 16 MRI studies (including structural and functional MRI), that used traditional machine learning methods and deep learning algorithms to automatically detect BD. The reported accuracies for EEG studies is about 90%, while the reported accuracies for MRI studies remains below the minimum level for clinical relevance, i.e. about 80% of the classification outcome for traditional machine learning methods. However, deep learning techniques have generally achieved accuracies higher than 95%. Conclusion: Research utilizing machine learning applied to EEG signals and brain images has provided proof of concept for how this innovative technique can help psychiatrists distinguish BD patients from healthy people. However, the results have been somewhat contradictory and we must keep away from excessive optimistic interpretations of the findings. Much progress is still needed to reach the level of clinical practice in this field.

Prediction Models for Suicide Attempts among Adolescents Using Machine Learning Techniques

OBJECTIVE

Suicide attempts (SAs) in adolescents are difficult to predict although it is a leading cause of death among adolescents. This study aimed to develop and evaluate SA prediction models based on six different machine learning (ML) algorithms for Korean adolescents using data from online surveys.

METHODS

Data were extracted from the 2011-2018 Korea Youth Risk Behavior Survey (KYRBS), an ongoing annual national survey. The participants comprised 468,482 nationally representative adolescents from 400 middle and 400 high schools, aged 12 to 18. The models were trained using several classic ML methods and then tested on internal and external independent datasets; performance metrics were calculated. Data analysis was performed from March 2020 to June 2020.

RESULTS

Among the 468,482 adolescents included in the analysis, 15,012 cases (3.2%) were identified as having made an SA. Three features (suicidal ideation, suicide planning, and grade) were identified as the most important predictors. The performance of the six ML models on the internal testing dataset was good, with both the area under the receiver operating characteristic curve (AUROC) and area under the precision-recall curve (AUPRC) ranging from 0.92 to 0.94. Although the AUROC of all models on the external testing dataset (2018 KYRBS) ranged from 0.93 to 0.95, the AUPRC of the models was approximately 0.5.

CONCLUSION

The developed and validated SA prediction models can be applied to detect high risks of SA. This approach could facilitate early intervention in the suicide crisis and may ultimately contribute to suicide prevention for adolescents.

Prediction of suicide attempts in a prospective cohort study with a nationally representative sample of the US population

BACKGROUND

There is still little knowledge of objective suicide risk stratification.

METHODS

This study aims to develop models using machine-learning approaches to predict suicide attempt (1) among survey participants in a nationally representative sample and (2) among participants with lifetime major depressive episodes. We used a cohort called the National Epidemiologic Survey on Alcohol and Related Conditions (NESARC) that was conducted in two waves and included a nationally representative sample of the adult population in the United States. Wave 1 involved 43 093 respondents and wave 2 involved 34 653 completed face-to-face reinterviews with wave 1 participants. Predictor variables included clinical, stressful life events, and sociodemographic variables from wave 1; outcome included suicide attempt between wave 1 and wave 2.

RESULTS

The model built with elastic net regularization distinguished individuals who had attempted suicide from those who had not with an area under the ROC curve (AUC) of 0.89, balanced accuracy 81.86%, specificity 89.22%, and sensitivity 74.51% for the general population. For participants with lifetime major depressive episodes, AUC was 0.89, balanced accuracy 81.64%, specificity 85.86%, and sensitivity 77.42%. The most important predictor variables were a diagnosis of borderline personality disorder, post-traumatic stress disorder, and being of Asian descent for the model in all participants; and previous suicide attempt, borderline personality disorder, and overnight stay in hospital because of depressive symptoms for the model in participants with lifetime major depressive episodes. Random forest and artificial neural networks had similar performance.

CONCLUSIONS

Risk for suicide attempt can be estimated with high accuracy.

Integrating biobehavioral information to predict mood disorder suicide risk

The will to live and the ability to maintain one's well-being are crucial for survival. Yet, almost a million people die by suicide globally each year (Aleman and Denys, 2014), making premature deaths due to suicide a significant public health problem (Saxena et al., 2013). The expression of suicidal behaviors is a complex phenotype with documented biological, psychological, clinical, and sociocultural risk factors (Turecki et al., 2019). From a brain disease perspective, suicide is associated with neuroanatomical, neurophysiological, and neurochemical dysregulations of brain networks involved in integrating and contextualizing cognitive and emotional regulatory behaviors. From a symptom perspective, diagnostic measures of dysregulated mood states like major depressive symptoms are associated with over sixty percent of suicide deaths worldwide (Saxena et al., 2013). This paper reviews the neurobiological and clinical phenotypic correlates for mood dysregulations and suicidal phenotypes. We further propose machine learning approaches to integrate neurobiological measures with dysregulated mood symptoms to elucidate the role of inflammatory processes as neurobiological risk factors for suicide.

Effectiveness of Artificial Intelligence Methods in Personalized Aggression Risk Prediction within Inpatient Psychiatric Treatment Settings—A Systematic Review

Aggression risk assessments are vital to prevent injuries and morbidities amongst patients and staff in psychiatric settings. More recent studies have harnessed artificial intelligence (AI) methods such as machine learning algorithms to determine factors associated with aggression in psychiatric treatment settings. In this review, using Cooper’s five-stage review framework, we aimed to evaluate the: (1) predictive accuracy, and (2) clinical variables associated with AI-based aggression risk prediction amongst psychiatric inpatients. Databases including PubMed, Cochrane, Scopus, PsycINFO, CINAHL were searched for relevant articles until April 2022. The eight included studies were independently evaluated using critical appraisal tools for systematic review developed by Joanna Briggs Institute. Most of the studies (87.5%) examined health records in predicting aggression and reported acceptable to excellent accuracy with specific machine learning algorithms employed (area under curve range 0.75–0.87). No particular machine learning algorithm outperformed the others consistently across studies (area under curve range 0.61–0.87). Relevant factors identified with aggression related to demographic and social profile, past aggression, forensic history, other psychiatric history, psychopathology, challenging behaviors and management domains. The limited extant studies have highlighted a potential role for the use of AI methods to clarify factors associated with aggression in psychiatric inpatient treatment settings.

Machine learning prediction of suicidal ideation, planning, and attempt among Korean adults: A population-based study

Background

Suicide remains the leading cause of premature death in South Korea. This study aims to develop machine learning algorithms for screening Korean adults at risk for suicidal ideation and suicide planning or attempt.

Methods

Two sets of balanced data for Korean adults aged 19–64 years were drawn from the 2012–2019 waves of the Korea Welfare Panel Study using the random down-sampling method (N = 3292 for the prediction of suicidal ideation, N = 488 for the prediction of suicide planning or attempt). Demographic, socioeconomic, and psychosocial characteristics were used to predict suicidal ideation and suicide planning or attempt. Four machine-learning classifiers (logistic regression, random forest, support vector machine, and extreme gradient boosting) were tuned and cross-validated.

Results

All four algorithms demonstrated satisfactory classification performance in predicting suicidal ideation (sensitivity 0.808–0.853, accuracy 0.843–0.863) and suicide planning or attempt (sensitivity 0.814–0.861, accuracy 0.864–0.884). Extreme gradient boosting was the best-performing algorithm for predicting both suicidal outcomes. The most important predictors were depressive symptoms, self-esteem, income, consumption, and life satisfaction. The algorithms trained with the top two predictors, depressive symptoms and self-esteem, showed comparable classification performance in predicting suicidal ideation (sensitivity 0.801–0.839, accuracy 0.841–0.846) and suicide planning or attempt (sensitivity 0.814–0.837, accuracy 0.874–0.884).

Limitations

Suicidal ideation and behaviors may be under-reported due to social desirability bias. Causality is not established.

Discussion

More than 80% of individuals at risk for suicidal ideation and suicide planning or attempt could be predicted by a number of mental and socioeconomic characteristics of respondents. This finding suggests the potential of developing a quick screening tool based on the known risk factors and applying it to primary care or community settings for early intervention.

•

This study develops machine learning models to predict suicidal ideation and behaviors.

•

Logistic regression, random forest, support vector machine, and extreme gradient boosting are used.

•

The algorithms correctly identifyed 80–90% of suicidal cases.

•

The algorithms with the top two predictors (depressive symptoms and self-esteem) could achieve comparable accuracy.

•

Our findings can be used to design a quick screening tool for use in primary care or community settings.

Social cognition and bipolar disorder: pending questions and unexplored topics

Social cognition has gained prominence in psychiatric research, beginning with schizophrenia and more recently in bipolar disorder. Considering the relevance of this domain to interpersonal relationships and functionality, we aimed to explore the fundamental research and clinical issues regarding social cognition and discuss future directions and challenges in the field of bipolar disorder.

Structured data vs. unstructured data in machine learning prediction models for suicidal behaviors: A systematic review and meta-analysis

Suicide remains a leading cause of preventable death worldwide, despite advances in research and decreases in mental health stigma through government health campaigns. Machine learning (ML), a type of artificial intelligence (AI), is the use of algorithms to simulate and imitate human cognition. Given the lack of improvement in clinician-based suicide prediction over time, advancements in technology have allowed for novel approaches to predicting suicide risk. This systematic review and meta-analysis aimed to synthesize current research regarding data sources in ML prediction of suicide risk, incorporating and comparing outcomes between structured data (human interpretable such as psychometric instruments) and unstructured data (only machine interpretable such as electronic health records). Online databases and gray literature were searched for studies relating to ML and suicide risk prediction. There were 31 eligible studies. The outcome for all studies combined was AUC = 0.860, structured data showed AUC = 0.873, and unstructured data was calculated at AUC = 0.866. There was substantial heterogeneity between the studies, the sources of which were unable to be defined. The studies showed good accuracy levels in the prediction of suicide risk behavior overall. Structured data and unstructured data also showed similar outcome accuracy according to meta-analysis, despite different volumes and types of input data.

Identifying posttraumatic stress disorder staging from clinical and sociodemographic features: a proof-of-concept study using a machine learning approach

This proof-of-concept study aimed to investigate the viability of a predictive model to support posttraumatic stress disorder (PTSD) staging. We performed a naturalistic, cross-sectional study at two Brazilian centers: the Psychological Trauma Research and Treatment (NET-Trauma) Program at Universidade Federal of Rio Grande do Sul, and the Program for Research and Care on Violence and PTSD (PROVE), at Universidade Federal of São Paulo. Five supervised machine-learning algorithms were tested: Elastic Net, Gradient Boosting Machine, Random Forest, Support Vector Machine, and C5.0, using clinical (Clinician-Administered PTSD Scale version 5) and sociodemographic features. A hundred and twelve patients were enrolled (61 from NET-Trauma and 51 from PROVE). We found a model with four classes suitable for the PTSD staging, with best performance metrics using the C5.0 algorithm to CAPS-5 15-items plus sociodemographic features, with an accuracy of 65.6% for the train dataset and 52.9% for the test dataset (both significant). The number of symptoms, CAPS-5 total score, global severity score, and presence of current/previous trauma events appear as main features to predict PTSD staging. This is the first study to evaluate staging in PTSD with machine learning algorithms using accessible clinical and sociodemographic features, which may be used in future research.

Artificial Intelligence Tools for Suicide Prevention in Adolescents and Young Adults

Background:

Artificial Intelligence is making a significant transformation in human lives. Its application in the medical and healthcare field has been also observed making an impact and improving overall outcomes. There has been a quest for similar processes in mental health due to the lack of observable changes in the areas of suicide prevention. In the last five years, there has been an emerging body of empirical research applying the technology of artificial intelligence (AI) and machine learning (ML) in mental health.

Objective:

To review the clinical applicability of the AI/ML-based tools in suicide prevention.

Methods:

The compelling question of predicting suicidality has been the focus of this research. We performed a broad literature search and then identified 36 articles relevant to meet the objectives of this review. We review the available evidence and provide a brief overview of the advances in this field.

Conclusion:

In the last five years, there has been more evidence supporting the implementation of these algorithms in clinical practice. Its current clinical utility is limited to using electronic health records and could be highly effective in conjunction with existing tools for suicide prevention. Other potential sources of relevant data include smart devices and social network sites. There are some serious questions about data privacy and ethics which need more attention while developing these new modalities in suicide research.

Predictors of suicide attempt in patients with obsessive-compulsive disorder: an exploratory study with machine learning analysis

BACKGROUND

Patients with obsessive-compulsive disorder (OCD) are at increased risk for suicide attempt (SA) compared to the general population. However, the significant risk factors for SA in this population remains unclear - whether these factors are associated with the disorder itself or related to extrinsic factors, such as comorbidities and sociodemographic variables. This study aimed to identify predictors of SA in OCD patients using a machine learning algorithm.

METHODS

A total of 959 outpatients with OCD were included. An elastic net model was performed to recognize the predictors of SA among OCD patients, using clinical and sociodemographic variables.

RESULTS

The prevalence of SA in our sample was 10.8%. Relevant predictors of SA founded by the elastic net algorithm were the following: previous suicide planning, previous suicide thoughts, lifetime depressive episode, and intermittent explosive disorder. Our elastic net model had a good performance and found an area under the curve of 0.95.

CONCLUSIONS

This is the first study to evaluate risk factors for SA among OCD patients using machine learning algorithms. Our results demonstrate an accurate risk algorithm can be created using clinical and sociodemographic variables. All aspects of suicidal phenomena need to be carefully investigated by clinicians in every evaluation of OCD patients. Particular attention should be given to comorbidity with depressive symptoms.

Identifying long-term and imminent suicide predictors in a general population and a clinical sample with machine learning

BACKGROUND

Machine learning (ML) is increasingly used to predict suicide deaths but their value for suicide prevention has not been established. Our first objective was to identify risk and protective factors in a general population. Our second objective was to identify factors indicating imminent suicide risk.

METHODS

We used survival and ML models to identify lifetime predictors using the Cohort of Norway (n=173,275) and hospital diagnoses in a Saskatoon clinical sample (n=12,614). The mean follow-up times were 17 years and 3 years for the Cohort of Norway and Saskatoon respectively. People in the clinical sample had a longitudinal record of hospital visits grouped in six-month intervals. We developed models in a training set and these models predicted survival probabilities in held-out test data.

RESULTS

In the general population, we found that a higher proportion of low-income residents in a county, mood symptoms, and daily smoking increased the risk of dying from suicide in both genders. In the clinical sample, the only predictors identified were male gender and older age.

CONCLUSION

Suicide prevention probably requires individual actions with governmental incentives. The prediction of imminent suicide remains highly challenging, but machine learning can identify early prevention targets.

A machine learning approach to identification of self-harm and suicidal ideation among military and police Veterans

LAY SUMMARY

Combat Veterans are vulnerable to suicidal thoughts and behaviour. Many who die by suicide deny having suicidal ideation (SI). Typically, researchers try to find variables indicating the presence of SI using traditional statistical approaches. These approaches do not possess the capacity to detect highly complex multivariable interactions. In contrast, machine learning (ML) is designed to detect such patterns and can consequently yield much higher predictive accuracy. In this study, the authors trained ML algorithms using 192 variables extracted from questionnaires administered to 738 Veterans and serving personnel to detect the presence of self-harm and SI (SHSI). Using the 10 most predictive non-suicide-related items, the ML algorithms could detect SHSI with 75.3% accuracy. Most of these items reflect psychological phenomena that can change quickly over time, allowing repeated risk reassessment from day to day. The study’s findings suggest that ML methods may play an important role in the discovery, within a large data set, of predictive patterns that might be useful in suicide risk assessment.

The Potential Impact of Adjunct Digital Tools and Technology to Help Distressed and Suicidal Men: An Integrative Review

Suicidal men feel the need to be self-reliant and that they cannot find another way out of relationship or socioeconomic issues. Suicide prevention is of crucial importance worldwide. The much higher rate of suicide in men engenders action. The prelude is a subjective experience that can be very isolating and severely distressing. Men may not realize a change in their thinking and behaviors, which makes it more difficult to seek and get help, thereby interrupting a "downward spiral". Stoicism often prevents men from admitting to their personal struggle. The lack of "quality" connections and "non-tailored" therapies has led to a high number of men "walking out" on traditional clinical approaches. But there are complicated relationships in motivations and formative behaviors of suicide with regards to emotional state, psychiatric disorders, interpersonal life events and suicidal behavior method selection. Middle-aged and older men have alternated as the most at-risk of suicide. There is no one solution that applies to all men, but digital tools may be of assistance (e.g., video conferences, social networks, telephone calls, and emails). Digital interventions require higher levels of effectiveness for distress and suicidality but self-guided approaches may be the most suitable for men especially where linked with an integrated online suicide prevention platform (e.g., quick response with online chats, phone calls, and emails). Furthermore, technology-enabled models of care offer promise to advance appropriate linking to mental health services through better and faster understanding of the specific needs of individuals (e.g., socio-cultural) and the type and level of suicidality experienced. Long-term evidence for suicidality and its evaluation may benefit from progressing human computer-interaction and providing impetus for an eminent integrated digital platform.

A comparative study of machine learning techniques for suicide attempts predictive model

Current suicide risk assessments for predicting suicide attempts are time consuming, of low predictive value and have inadequate reliability. This paper aims to develop a predictive model for suicide attempts among patients with depression using machine learning algorithms as well as presents a comparative study on single predictive models with ensemble predictive models for differentiating depressed patients with suicide attempts from non-suicide attempters. We applied and trained eight different machine learning algorithms using a dataset that consists of 75 patients diagnosed with a depressive disorder. A recursive feature elimination was used to reduce the features via three-fold cross validation. An ensemble predictive models outperformed the single predictive models. Voting and bagging revealed the highest accuracy of 92% compared to other machine learning algorithms. Our findings indicate that history of suicide attempt, religion, race, suicide ideation and severity of clinical depression are useful factors for prediction of suicide attempts.

Detecting suicidal risk using MMPI-2 based on machine learning algorithm

Minnesota Multiphasic Personality Inventory-2 (MMPI-2) is a widely used tool for early detection of psychological maladjustment and assessing the level of adaptation for a large group in clinical settings, schools, and corporations. This study aims to evaluate the utility of MMPI-2 in assessing suicidal risk using the results of MMPI-2 and suicidal risk evaluation. A total of 7,824 datasets collected from college students were analyzed. The MMPI-2-Resturcutred Clinical Scales (MMPI-2-RF) and the response results for each question of the Mini International Neuropsychiatric Interview (MINI) suicidality module were used. For statistical analysis, random forest and K-Nearest Neighbors (KNN) techniques were used with suicidal ideation and suicide attempt as dependent variables and 50 MMPI-2 scale scores as predictors. On applying the random forest method to suicidal ideation and suicidal attempts, the accuracy was 92.9% and 95%, respectively, and the Area Under the Curves (AUCs) were 0.844 and 0.851, respectively. When the KNN method was applied, the accuracy was 91.6% and 94.7%, respectively, and the AUCs were 0.722 and 0.639, respectively. The study confirmed that machine learning using MMPI-2 for a large group provides reliable accuracy in classifying and predicting the subject's suicidal ideation and past suicidal attempts.

Information and Communication Technology Use in Suicide Prevention: Scoping Review

BACKGROUND

The use of information and communication technology (ICT) in suicide prevention has progressed rapidly over the past decade. ICT plays a major role in suicide prevention, but research on best and promising practices has been slow.

OBJECTIVE

This paper aims to explore the existing literature on ICT use in suicide prevention to answer the following question: what are the best and most promising ICT practices for suicide prevention?

METHODS

A scoping search was conducted using the following databases: PubMed, PsycINFO, Sociological Abstracts, and IEEE Xplore. These databases were searched for articles published between January 1, 2013, and December 31, 2018. The five stages of the scoping review process were as follows: identifying research questions; targeting relevant studies; selecting studies; charting data; and collating, summarizing, and reporting the results. The World Health Organization suicide prevention model was used according to the continuum of universal, selective, and indicated prevention.

RESULTS

Of the 3848 studies identified, 115 (2.99%) were selected. Of these, 10 regarded the use of ICT in universal suicide prevention, 53 referred to the use of ICT in selective suicide prevention, and 52 dealt with the use of ICT in indicated suicide prevention.

CONCLUSIONS

The use of ICT plays a major role in suicide prevention, and many promising programs were identified through this scoping review. However, large-scale evaluation studies are needed to further examine the effectiveness of these programs and strategies. In addition, safety and ethics protocols for ICT-based interventions are recommended.

[Cross-sectoral therapeutic concepts and innovative technologies: new opportunities for the treatment of patients with mental disorders]

Mental disorders are widespread and a major public health problem. The risk of developing a mental disorder at some point in life is around 40%. Therefore, mental disorders are among the most common diseases. Despite the introduction of newer psychotropic drugs, disorder-specific psychotherapy and stimulation techniques, many of those affected still show insufficient symptom remission and a chronic course of the disorder. Conceptual and technological progress in recent years has enabled a new, more flexible and personalized form of mental health care. Both the traditional therapeutic concepts and newer decentralized, modularly structured, track units, together with innovative digital technologies, will offer individualized therapeutic options in order to alleviate symptoms and improve quality of life of patients with mental illnesses. The primary goal of closely combining inpatient care concepts with innovative technologies is to provide comprehensive therapy and aftercare concepts for all individual needs of patients with mental disorders. Last but not least, this also ensures that specialist psychiatric treatment is available regardless of location. In twenty-first century psychiatry, modern care structures must be effectively linked to the current dynamics of digital transformation. This narrative review is dedicated to the theoretical and practical aspects of a cross-sectoral treatment system combined with innovative digital technologies in the psychiatric-psychotherapeutic field. The authors aim to illuminate these therapy modalities using the example of the Central Institute of Mental Health in Mannheim.

Teasing out Artificial Intelligence in Medicine: An Ethical Critique of Artificial Intelligence and Machine Learning in Medicine

The rapid adoption and implementation of artificial intelligence in medicine creates an ontologically distinct situation from prior care models. There are both potential advantages and disadvantages with such technology in advancing the interests of patients, with resultant ontological and epistemic concerns for physicians and patients relating to the instatiation of AI as a dependent, semi- or fully-autonomous agent in the encounter. The concept of libertarian paternalism potentially exercised by AI (and those who control it) has created challenges to conventional assessments of patient and physician autonomy. The unclear legal relationship between AI and its users cannot be settled presently, an progress in AI and its implementation in patient care will necessitate an iterative discourse to preserve humanitarian concerns in future models of care. This paper proposes that physicians should neither uncritically accept nor unreasonably resist developments in AI but must actively engage and contribute to the discourse, since AI will affect their roles and the nature of their work. One's moral imaginative capacity must be engaged in the questions of beneficence, autonomy, and justice of AI and whether its integration in healthcare has the potential to augment or interfere with the ends of medical practice.

Psychotic symptoms during bipolar depressive episodes and suicidal ideation

BACKGROUND

Psychotic symptoms during bipolar depressive episodes, especially in outpatients, are under recognized and studied by clinicians and researchers. We examined the relationship between psychotic symptoms during a depressive episode and suicidal ideation in bipolar patients.

METHODS

Participants (N = 351) were adult, depressed outpatients with bipolar disorder (BD) in a comparative effectiveness study of quetiapine versus lithium. Psychotic symptoms were assessed via Bipolar Inventory of Signs and Symptoms Scale (BISS) and depressive episodes via Mini-International Neuropsychiatric Interview (MINI). Because only 4.84% (N = 17) endorsed psychotic symptoms, we performed iterative multivariate matching with non-psychotic participants. On every matched population, a multiple regression analysis examined whether psychotic symptoms were associated with suicidal ideation, via the Concise Health Risk Taking scale (CHRT-12).

RESULTS

Averaged across the 50 matched populations, current psychotic symptoms predicted active suicidal ideation on the CHRT, but not a passive propensity toward suicide or total CHRT scores, after adjusting for common correlates of suicidality (e.g., previous suicidal behavior) (β=0.59, p=.01, R2= 0.41).

LIMITATIONS

Our study was limited by three factors. First, the generalizability of our study was limited as the sample included only outpatients. Next, the analysis was cross-sectional and does not allow for causal interpretation. Lastly, our study lacked information regarding the content and mood congruency of participants' psychosis.

CONCLUSION

While a small proportion of BD outpatients had current symptoms of psychosis during their depressive episode, those who did were more likely to endorse active suicidal thoughts, including suicide methods and plans.

Hybrid bag of approaches to characterize selection criteria for cohort identification

OBJECTIVE

The 2018 National NLP Clinical Challenge (2018 n2c2) focused on the task of cohort selection for clinical trials, where participating systems were tasked with analyzing longitudinal patient records to determine if the patients met or did not meet any of the 13 selection criteria. This article describes our participation in this shared task.

MATERIALS AND METHODS

We followed a hybrid approach combining pattern-based, knowledge-intensive, and feature weighting techniques. After preprocessing the notes using publicly available natural language processing tools, we developed individual criterion-specific components that relied on collecting knowledge resources relevant for these criteria and pattern-based and weighting approaches to identify "met" and "not met" cases.

RESULTS

As part of the 2018 n2c2 challenge, 3 runs were submitted. The overall micro-averaged F1 on the training set was 0.9444. On the test set, the micro-averaged F1 for the 3 submitted runs were 0.9075, 0.9065, and 0.9056. The best run was placed second in the overall challenge and all 3 runs were statistically similar to the top-ranked system. A reimplemented system achieved the best overall F1 of 0.9111 on the test set.

DISCUSSION

We highlight the need for a focused resource-intensive effort to address the class imbalance in the cohort selection identification task.

CONCLUSION

Our hybrid approach was able to identify all selection criteria with high F1 performance on both training and test sets. Based on our participation in the 2018 n2c2 task, we conclude that there is merit in continuing a focused criterion-specific analysis and developing appropriate knowledge resources to build a quality cohort selection system.

Screening for Depression in Mobile Devices Using Patient Health Questionnaire-9 (PHQ-9) Data: A Diagnostic Meta-Analysis via Machine Learning Methods

PURPOSE

Depression is a symptom commonly encountered in primary care; however, it is often not detected by doctors. Recently, disease diagnosis and treatment approaches have been attempted using smart devices. In this study, instrumental effectiveness was confirmed with the diagnostic meta-analysis of studies that demonstrated the diagnostic effectiveness of PHQ-9 for depression using mobile devices.

PATIENTS AND METHODS

We found all published and unpublished studies through EMBASE, MEDLINE, MEDLINE In-Process, and PsychINFO up to March 26, 2021. We performed a meta-analysis by including 1099 subjects in four studies. We performed a diagnostic meta-analysis according to the PHQ-9 cut-off score and machine learning algorithm techniques. Quality assessment was conducted using the QUADAS-2 tool. Data on the sensitivity and specificity of the studies included in the meta-analysis were extracted in a standardized format. Bivariate and summary receiver operating characteristic (SROC) curve were constructed using the metandi, midas, metabias, and metareg functions of the Stata algorithm meta-analysis words.

RESULTS

Using four studies out of the 5476 papers searched, a diagnostic meta-analysis of the PHQ-9 scores of 1099 people diagnosed with depression was performed. The pooled sensitivity and specificity were 0.797 (95% CI = 0.642-0.895) and 0.85 (95% CI = 0.780-0.900), respectively. The diagnostic odds ratio was 22.16 (95% CI = 7.273-67.499). Overall, a good balance was maintained, and no heterogeneity or publication bias was presented.

CONCLUSION

Through various machine learning algorithm techniques, it was possible to confirm that PHQ-9 depression screening in mobiles is an effective diagnostic tool when integrated into a diagnostic meta-analysis.

Identification of Suicidal Ideation in the Canadian Community Health Survey-Mental Health Component Using Deep Learning

Introduction: Suicidal ideation (SI) is prevalent in the general population, and is a risk factor for suicide. Predicting which patients are likely to have SI remains challenging. Deep Learning (DL) may be a useful tool in this context, as it can be used to find patterns in complex, heterogeneous, and incomplete datasets. An automated screening system for SI could help prompt clinicians to be more attentive to patients at risk for suicide. Methods: Using the Canadian Community Health Survey-Mental Health Component, we trained a DL model based on 23,859 survey responses to classify patients with and without SI. Models were created to classify both lifetime SI and SI over the last 12 months. From 582 possible parameters we produced 96- and 21-feature versions of the models. Models were trained using an undersampling procedure that balanced the training set between SI and non-SI; validation was done on held-out data. Results: For lifetime SI, the 96 feature model had an Area under the receiver operating curve (AUC) of 0.79 and the 21 feature model had an AUC of 0.77. For SI in the last 12 months the 96 feature model had an AUC of 0.71 and the 21 feature model had an AUC of 0.68. In addition, sensitivity analyses demonstrated feature relationships in line with existing literature. Discussion: Although further study is required to ensure clinical relevance and sample generalizability, this study is an initial proof of concept for the use of DL to improve identification of SI. Sensitivity analyses can help improve the interpretability of DL models. This kind of model would help start conversations with patients which could lead to improved care and a reduction in suicidal behavior.

Early identification of bipolar disorder among young adults - a 22-year community birth cohort

OBJECTIVE

We set forth to build a prediction model of individuals who would develop bipolar disorder (BD) using machine learning techniques in a large birth cohort.

METHODS

A total of 3748 subjects were studied at birth, 11, 15, 18, and 22 years of age in a community birth cohort. We used the elastic net algorithm with 10-fold cross-validation to predict which individuals would develop BD at endpoint (22 years) at each follow-up visit before diagnosis (from birth up to 18 years). Afterward, we used the best model to calculate the subgroups of subjects at higher and lower risk of developing BD and analyzed the clinical differences among them.

RESULTS

A total of 107 (2.8%) individuals within the cohort presented with BD type I, 26 (0.6%) with BD type II, and 87 (2.3%) with BD not otherwise specified. Frequency of female individuals was 58.82% (n = 150) in the BD sample and 53.02% (n = 1868) among the unaffected population. The model with variables assessed at the 18-year follow-up visit achieved the best performance: AUC 0.82 (CI 0.75-0.88), balanced accuracy 0.75, sensitivity 0.72, and specificity 0.77. The most important variables to detect BD at the 18-year follow-up visit were suicide risk, generalized anxiety disorder, parental physical abuse, and financial problems. Additionally, the high-risk subgroup of BD showed a high frequency of drug use and depressive symptoms.

CONCLUSIONS

We developed a risk calculator for BD incorporating both demographic and clinical variables from a 22-year birth cohort. Our findings support previous studies in high-risk samples showing the significance of suicide risk and generalized anxiety disorder prior to the onset of BD, and highlight the role of social factors and adverse life events.

Accurate Diagnosis of Suicide Ideation/Behavior Using Robust Ensemble Machine Learning: A University Student Population in the Middle East and North Africa (MENA) Region

Suicide is one of the most critical public health concerns in the world and the second cause of death among young people in many countries. However, to date, no study can diagnose suicide ideation/behavior among university students in the Middle East and North Africa (MENA) region using a machine learning approach. Therefore, stability feature selection and stacked ensembled decision trees were employed in this classification problem. A total of 573 university students responded to a battery of questionnaires. Three-fold cross-validation with a variety of performance indices was sued. The proposed diagnostic system had excellent balanced diagnosis accuracy (AUC = 0.90 [CI 95%: 0.86-0.93]) with a high correlation between predicted and observed class labels, fair discriminant power, and excellent class labeling agreement rate. Results showed that 23 items out of all items could accurately diagnose suicide ideation/behavior. These items were psychological problems and how to experience trauma, from the demographic variables, nine items from Post-Traumatic Stress Disorder Checklist (PCL-5), two items from Post Traumatic Growth (PTG), two items from the Patient Health Questionnaire (PHQ), six items from the Positive Mental Health (PMH) questionnaire, and one item related to social support. Such features could be used as a screening tool to identify young adults who are at risk of suicide ideation/behavior.

Exploring mood symptoms overlap in PTSD diagnosis: ICD-11 and DSM-5 criteria compared in a sample of subjects with Bipolar Disorder

BACKGROUND

The latest edition of the ICD (ICD-11) introduced relevant modifications to Post-traumatic Stress Disorder (PTSD) diagnostic criteria with respect to those of the DSM-5, including the exclusion of DSM-5 symptoms that potentially overlapped with mood disorders. To date, no study has yet investigated the differences in PTSD and its related symptoms, according to the two diagnostic systems in subjects with mood disorders. The aim of the present study was to compare the DSM-5 and ICD-11 diagnostic criteria for PTSD in a sample of patients with Bipolar Disorder (BD).

METHODS

An overall sample of 210 in-patients with BD completed the Trauma and Loss Spectrum-Self Report, assessing post-traumatic stress symptoms, to compare symptomatological PTSD diagnosis according to either the DSM-5 or the ICD-11 criteria.

RESULTS

DSM-5 PTSD was detected in 41% of the whole sample, whereas ICD-11 PTSD in 31.8%. The two diagnostic systems showed good concordance (Cohen's k = 0.643), whereas the concordance of re-experiencing and arousal criteria were moderate (Cohen's k = 0.578) and good (Cohen's k = 0.791), respectively. Almost all the subjects with a diagnosis of ICD-11 PTSD (92.5%) endorsed the "negative alterations in cognitions and mood" DSM-5 criterion.

LIMITATIONS

The small size, the use of a self-report instrument.

CONCLUSION

Our findings show high rates of PTSD and post-traumatic stress symptoms among subjects with BD according to both DSM-5 and ICD-11 criteria, despite significantly lower with the latter. However, potentially DSM-5 mood overlapping symptoms appear to be significantly higher among bipolar patients with ICD-11 PTSD with respect to those without.

Prediction of Suicide-Related Events by Analyzing Electronic Medical Records from PTSD Patients with Bipolar Disorder

Around 800,000 people worldwide die from suicide every year and it’s the 10th leading cause of death in the US. It is of great value to build a mathematic model that can accurately predict suicide especially in high-risk populations. Several different ML-based models were trained and evaluated using features obtained from electronic medical records (EMRs). The contribution of each feature was calculated to determine how it impacted the model predictions. The best-performing model was selected for analysis and decomposition. Random forest showed the best performance with true positive rates (TPR) and positive predictive values (PPV) of greater than 80%. The use of Sertraline, Fentanyl, Aripiprazole, Lamotrigine, and Tramadol were strong indicators for no SREs within one year. The use of Haloperidol, Trazodone and Citalopram, a diagnosis of autistic disorder, schizophrenic disorder, or substance use disorder at the time of a diagnosis of both PTSD and bipolar disorder, predicted the onset of SREs within one year. Additional features with potential protective or hazardous effects for SREs were identified by the model. We constructed an ML-based model that was successful in identifying patients in a subpopulation at high-risk for SREs within a year of diagnosis of both PTSD and bipolar disorder. The model also provides feature decompositions to guide mechanism studies. The validation of this model with additional EMR datasets will be of great value in resource allocation and clinical decision making.

Aberrant functional connectivity and graph properties in bipolar II disorder with suicide attempts

OBJECTIVE

The physiological mechanism of suicide attempt (SA) in bipolar II disorder (BD-II) remains only partially understood. The study seeks to identify the dysfunction pattern in suicide brain for BD-II patients.

METHODS

Graph theory was utilized to explore topological properties at whole-brain, module and region levels based on resting-state functional MRI (rs-fMRI) data, which acquired from 38 un-medicated BD-II patients with at least one SA, 60 none SA (NSA) patients and 69 healthy controls (HCs). Finally, the correlation relationship between graph metrics and clinical variables were estimated.

RESULTS

Compared with NSA patients and HCs, the functional connectivity strength between limbic/sub-cortical (LIMB/SubC) and frontoparietal network (FPN) were significantly weakened. Nodal strength in left head of caudate nucleus (HCN), raphe nucleus (RN), right nucleus accumbens (NAcc), right subgenual anterior cingulate cortex (sgACC) and nodal efficiency in right sgACC, right HCN for SA patients were significantly reduced relative to NSA and HCs. In particular, nodal strength in RN and nodal efficiency in right sgACC showed a significant negative correlation with Nurses' Global Assessment of Suicide Risk (NGASR) scores.

LIMITATIONS

This is a single-mode cross-sectional study, the results were not verified by multi-center data.

CONCLUSIONS

The abnormal disrupted FC between LIMB/SubC and FPN is associated with SA in BD-II patients, which increased the susceptibility of suicide. Especially, the dysfunction in RN and right sgACC predict a higher suicide risk in BD-II patients.The results can help us to understand the suicide mechanism and early judgment of suicidal behaviors for BD-II patients.

Exploring Suicidal Ideation Using an Innovative Mobile App-Strength Within Me: The Usability and Acceptability of Setting up a Trial Involving Mobile Technology and Mental Health Service Users

BACKGROUND

Suicide is a growing global public health problem that has resulted in an increase in the demand for psychological services to address mental health issues. It is expected that 1 in 6 people on a waiting list for mental health services will attempt suicide. Although suicidal ideation has been shown to be linked to a higher risk of death by suicide, not everybody openly discloses their suicidal thoughts or plans to friends and family or seeks professional help before suicide. Therefore, new methods are needed to track suicide risk in real time together with a better understanding of the ways in which people communicate or express their suicidality. Considering the dynamic nature and challenges in understanding suicide ideation and suicide risk, mobile apps could be better suited to prevent suicide as they have the ability to collect real-time data.

OBJECTIVE

This study aims to report the practicalities and acceptability of setting up and trialing digital technologies within an inpatient mental health setting in the United Kingdom and highlight their implications for future studies.

METHODS

Service users were recruited from 6 inpatient wards in the north west of England. Service users who were eligible to participate and provided consent were given an iPhone and Fitbit for 7 days and were asked to interact with a novel phone app, Strength Within Me (SWiM). Interaction with the app involved journaling (recording daily activities, how this made them feel, and rating their mood) and the option to create safety plans for emotions causing difficulties (identifying strategies that helped with these emotions). Participants also had the option to allow the study to access their personal Facebook account to monitor their social media use and activity. In addition, clinical data (ie, assessments conducted by trained researchers targeting suicidality, depression, and sleep) were also collected.

RESULTS

Overall, 43.0% (80/186 response rate) of eligible participants were recruited for the study. Of the total sample, 67 participants engaged in journaling, with the average number of entries per user being 8.2 (SD 8.7). Overall, only 24 participants created safety plans and the most common difficult emotion to be selected was feeling sad (n=21). This study reports on the engagement with the SWiM app, the technical difficulties the research team faced, the importance of building key relationships, and the implications of using Facebook as a source to detect suicidality.

CONCLUSIONS

To develop interventions that can be delivered in a timely manner, prediction of suicidality must be given priority. This paper has raised important issues and highlighted lessons learned from implementing a novel mobile app to detect the risk of suicidality for service users in an inpatient setting.

Artificial Intelligence and Suicide Prevention: A Systematic Review of Machine Learning Investigations

Suicide is a leading cause of death that defies prediction and challenges prevention efforts worldwide. Artificial intelligence (AI) and machine learning (ML) have emerged as a means of investigating large datasets to enhance risk detection. A systematic review of ML investigations evaluating suicidal behaviors was conducted using PubMed/MEDLINE, PsychInfo, Web-of-Science, and EMBASE, employing search strings and MeSH terms relevant to suicide and AI. Databases were supplemented by hand-search techniques and Google Scholar. Inclusion criteria: (1) journal article, available in English, (2) original investigation, (3) employment of AI/ML, (4) evaluation of a suicide risk outcome. N = 594 records were identified based on abstract search, and 25 hand-searched reports. N = 461 reports remained after duplicates were removed, n = 316 were excluded after abstract screening. Of n = 149 full-text articles assessed for eligibility, n = 87 were included for quantitative synthesis, grouped according to suicide behavior outcome. Reports varied widely in methodology and outcomes. Results suggest high levels of risk classification accuracy (>90%) and Area Under the Curve (AUC) in the prediction of suicidal behaviors. We report key findings and central limitations in the use of AI/ML frameworks to guide additional research, which hold the potential to impact suicide on broad scale.

Testing Suicide Risk Prediction Algorithms Using Phone Measurements With Patients in Acute Mental Health Settings: Feasibility Study

BACKGROUND

Digital phenotyping and machine learning are currently being used to augment or even replace traditional analytic procedures in many domains, including health care. Given the heavy reliance on smartphones and mobile devices around the world, this readily available source of data is an important and highly underutilized source that has the potential to improve mental health risk prediction and prevention and advance mental health globally.

OBJECTIVE

This study aimed to apply machine learning in an acute mental health setting for suicide risk prediction. This study uses a nascent approach, adding to existing knowledge by using data collected through a smartphone in place of clinical data, which have typically been collected from health care records.

METHODS

We created a smartphone app called Strength Within Me, which was linked to Fitbit, Apple Health kit, and Facebook, to collect salient clinical information such as sleep behavior and mood, step frequency and count, and engagement patterns with the phone from a cohort of inpatients with acute mental health (n=66). In addition, clinical research interviews were used to assess mood, sleep, and suicide risk. Multiple machine learning algorithms were tested to determine the best fit.

RESULTS

K-nearest neighbors (KNN; k=2) with uniform weighting and the Euclidean distance metric emerged as the most promising algorithm, with 68% mean accuracy (averaged over 10,000 simulations of splitting the training and testing data via 10-fold cross-validation) and an average area under the curve of 0.65. We applied a combined 5×2 F test to test the model performance of KNN against the baseline classifier that guesses training majority, random forest, support vector machine and logistic regression, and achieved F statistics of 10.7 (P=.009) and 17.6 (P=.003) for training majority and random forest, respectively, rejecting the null of performance being the same. Therefore, we have taken the first steps in prototyping a system that could continuously and accurately assess the risk of suicide via mobile devices.

CONCLUSIONS

Predicting for suicidality is an underaddressed area of research to which this paper makes a useful contribution. This is part of the first generation of studies to suggest that it is feasible to utilize smartphone-generated user input and passive sensor data to generate a risk algorithm among inpatients at suicide risk. The model reveals fair concordance between phone-derived and research-generated clinical data, and with iterative development, it has the potential for accurate discriminant risk prediction. However, although full automation and independence of clinical judgment or input would be a worthy development for those individuals who are less likely to access specialist mental health services, and for providing a timely response in a crisis situation, the ethical and legal implications of such advances in the field of psychiatry need to be acknowledged.

Early Detection of Heart Failure With Reduced Ejection Fraction Using Perioperative Data Among Noncardiac Surgical Patients: A Machine-Learning Approach

BACKGROUND

Heart failure with reduced ejection fraction (HFrEF) is a condition imposing significant health care burden. Given its syndromic nature and often insidious onset, the diagnosis may not be made until clinical manifestations prompt further evaluation. Detecting HFrEF in precursor stages could allow for early initiation of treatments to modify disease progression. Granular data collected during the perioperative period may represent an underutilized method for improving the diagnosis of HFrEF. We hypothesized that patients ultimately diagnosed with HFrEF following surgery can be identified via machine-learning approaches using pre- and intraoperative data.

METHODS

Perioperative data were reviewed from adult patients undergoing general anesthesia for major surgical procedures at an academic quaternary care center between 2010 and 2016. Patients with known HFrEF, heart failure with preserved ejection fraction, preoperative critical illness, or undergoing cardiac, cardiology, or electrophysiologic procedures were excluded. Patients were classified as healthy controls or undiagnosed HFrEF. Undiagnosed HFrEF was defined as lacking a HFrEF diagnosis preoperatively but establishing a diagnosis within 730 days postoperatively. Undiagnosed HFrEF patients were adjudicated by expert clinician review, excluding cases for which HFrEF was secondary to a perioperative triggering event, or any event not associated with HFrEF natural disease progression. Machine-learning models, including L1 regularized logistic regression, random forest, and extreme gradient boosting were developed to detect undiagnosed HFrEF, using perioperative data including 628 preoperative and 1195 intraoperative features. Training/validation and test datasets were used with parameter tuning. Test set model performance was evaluated using area under the receiver operating characteristic curve (AUROC), positive predictive value, and other standard metrics.

RESULTS

Among 67,697 cases analyzed, 279 (0.41%) patients had undiagnosed HFrEF. The AUROC for the logistic regression model was 0.869 (95% confidence interval, 0.829-0.911), 0.872 (0.836-0.909) for the random forest model, and 0.873 (0.833-0.913) for the extreme gradient boosting model. The corresponding positive predictive values were 1.69% (1.06%-2.32%), 1.42% (0.85%-1.98%), and 1.78% (1.15%-2.40%), respectively.

CONCLUSIONS

Machine-learning models leveraging perioperative data can detect undiagnosed HFrEF with good performance. However, the low prevalence of the disease results in a low positive predictive value, and for clinically meaningful sensitivity thresholds to be actionable, confirmatory testing with high specificity (eg, echocardiography or cardiac biomarkers) would be required following model detection. Future studies are necessary to externally validate algorithm performance at additional centers and explore the feasibility of embedding algorithms into the perioperative electronic health record for clinician use in real time.

Will machine learning applied to neuroimaging in bipolar disorder help the clinician? A critical review and methodological suggestions

OBJECTIVES

The existence of anatomofunctional brain abnormalities in bipolar disorder (BD) is now well established by magnetic resonance imaging (MRI) studies. To create diagnostic and prognostic tools, as well as identifying biologically valid subtypes of BD, research has recently turned towards the use of machine learning (ML) techniques. We assessed both supervised ML and unsupervised ML studies in BD to evaluate their robustness, reproducibility and the potential need for improvement.

METHOD

We systematically searched for studies using ML algorithms based on MRI data of patients with BD until February 2019.

RESULT

We identified 47 studies, 45 using supervised ML techniques and 2 including unsupervised ML analyses. Among supervised studies, 43 focused on diagnostic classification. The reported accuracies for classification of BD ranged between (a) 57% and 100%, for BD vs healthy controls; (b) 49.5% and 93.1% for BD vs patients with major depressive disorder; and (c) 50% and 96.2% for BD vs patients with schizophrenia. Reported accuracies for discriminating subjects genetically at risk for BD (either from control or from patients with BD) ranged between 64.3% and 88.93%.

CONCLUSIONS

Although there are strong methodological limitations in previous studies and an important need for replication in large multicentric samples, the conclusions of our review bring hope of future computer-aided diagnosis of BD and pave the way for other applications, such as treatment response prediction. To reinforce the reliability of future results we provide methodological suggestions for good practice in conducting and reporting MRI-based ML studies in BD.

Brain-derived neurotrophic factor blood levels after electroconvulsive therapy in patients with major depressive disorder: A systematic review and meta-analysis

Some evidence pointed out that Electro-Convulsive Treatment (ECT) could increase the level of brain-derived neurotrophic factor (BDNF) in depressive patients. However, there are some disagreements. The purpose of the study is through a systematic review and meta-analysis to evaluate BDNF levels after ECT in patients with Major depressive disorder. Two independent researchers searched of published articles in the databases of Cochrane Library, PubMed, MEDLINE, EMBASE and WanFang Data, from January 1990 to March 2019. The following key words were used: "depression" or "depressive disorder", "major depressive disorder", "unipolar depression", "brain-derived neurotrophic factor" or "BDNF", and "electroconvulsive" or "ECT". A total of 22 studies met the inclusion criteria of the meta-analysis and included into our analysis. BDNF levels were increased among patients with MDD after ECT (P = 0.000) in plasma samples. The standardized mean difference (SMD) was 0.695 (95 % CI: 0.402-0.988). We also found BDNF levels increased on one week and one month after finishing ECT (SMD = 0.491, 95 %CI: 0.150,0.833, P = 0.005; and SMD = 0.812, 95 %CI: 0.326,1.298, P = 0.001, respectively). Our findings suggest that BDNF levels may increase after ECT and may possibly be used as an indicator of treatment response after one or more weeks of ECT in patients with depression. However, additional investigation of BDNF levels with different ECT durations are needed in responders and non-responders.

Suicide and suicide attempts in adults: exploring suicide risk 24 months after a psychiatric emergency room visit

Objective:

Suicide risk (including attempted and completed suicide) should be measured over short periods of time after contacting health services. The objective of this study was to identify the patterns of attempted and completed suicides within 24-months of a psychiatric emergency department visit, as well as to investigate predictive risk factors, including sociodemographic and clinical variables, previous suicidal behavior, and service utilization.

Method:

A convenience sample (n=147), recruited at a general hospital’s psychiatric emergency room, included patients with suicidal ideation, suicidal plans or previous suicide attempts. These patients were followed for 24 months, focusing on two main outcomes: attempted and completed suicides.

Results:

After six months there were no completed suicides and 36 suicide attempts, while after 24 months there were seven completed suicides and 69 suicide attempts. A final logistic regression model for suicide attempts at 24 months identified somatic pathology and the number of previous psychiatric hospitalizations as predictive factors, with a good area under the receiver operating characteristic curve.

Conclusions:

The findings showed distinct patterns of attempted and completed suicides over time, indicating the importance of a systematic multidisciplinary suicide risk evaluation in psychiatric emergency rooms.