Machine learning methods to predict child posttraumatic stress: a proof of concept study

Multi-omics approaches in psychoneuroimmunology and health research: Conceptual considerations and methodological recommendations

The field of psychoneuroimmunology (PNI) has grown substantially in both relevance and prominence over the past 40 years. Notwithstanding its impressive trajectory, a majority of PNI studies are still based on a relatively small number of analytes. To advance this work, we suggest that PNI, and health research in general, can benefit greatly from adopting a multi-omics approach, which involves integrating data across multiple biological levels (e.g., the genome, proteome, transcriptome, metabolome, lipidome, and microbiome/metagenome) to more comprehensively profile biological functions and relate these profiles to clinical and behavioral outcomes. To assist investigators in this endeavor, we provide an overview of multi-omics research, highlight recent landmark multi-omics studies investigating human health and disease risk, and discuss how multi-omics can be applied to better elucidate links between psychological, nervous system, and immune system activity. In doing so, we describe how to design high-quality multi-omics studies, decide which biological samples (e.g., blood, stool, urine, saliva, solid tissue) are most relevant, incorporate behavioral and wearable sensing data into multi-omics research, and understand key data quality, integration, analysis, and interpretation issues. PNI researchers are addressing some of the most interesting and important questions at the intersection of psychology, neuroscience, and immunology. Applying a multi-omics approach to this work will greatly expand the horizon of what is possible in PNI and has the potential to revolutionize our understanding of mind-body medicine.

Predicting individual cases of major adolescent psychiatric conditions with artificial intelligence

Three-quarters of lifetime mental illness occurs by the age of 24, but relatively little is known about how to robustly identify youth at risk to target intervention efforts known to improve outcomes. Barriers to knowledge have included obtaining robust predictions while simultaneously analyzing large numbers of different types of candidate predictors. In a new, large, transdiagnostic youth sample and multidomain high-dimension data, we used 160 candidate predictors encompassing neural, prenatal, developmental, physiologic, sociocultural, environmental, emotional and cognitive features and leveraged three different machine learning algorithms optimized with a novel artificial intelligence meta-learning technique to predict individual cases of anxiety, depression, attention deficit, disruptive behaviors and post-traumatic stress. Our models tested well in unseen, held-out data (AUC ≥ 0.94). By utilizing a large-scale design and advanced computational approaches, we were able to compare the relative predictive ability of neural versus psychosocial features in a principled manner and found that psychosocial features consistently outperformed neural metrics in their relative ability to deliver robust predictions of individual cases. We found that deep learning with artificial neural networks and tree-based learning with XGBoost outperformed logistic regression with ElasticNet, supporting the conceptualization of mental illnesses as multifactorial disease processes with non-linear relationships among predictors that can be robustly modeled with computational psychiatry techniques. To our knowledge, this is the first study to test the relative predictive ability of these gold-standard algorithms from different classes across multiple mental health conditions in youth within the same study design in multidomain data utilizing >100 candidate predictors. Further research is suggested to explore these findings in longitudinal data and validate results in an external dataset.

Pediatrics in Artificial Intelligence Era: A Systematic Review on Challenges, Opportunities, and Explainability

BACKGROUND

The emergence of artificial intelligence (AI) tools such as ChatGPT and Bard is disrupting a broad swathe of fields, including medicine. In pediatric medicine, AI is also increasingly being used across multiple subspecialties. However, the practical application of AI still faces a number of key challenges. Consequently, there is a requirement for a concise overview of the roles of AI across the multiple domains of pediatric medicine, which the current study seeks to address.

AIM

To systematically assess the challenges, opportunities, and explainability of AI in pediatric medicine.

METHODOLOGY

A systematic search was carried out on peer-reviewed databases, PubMed Central, Europe PubMed Central, and grey literature using search terms related to machine learning (ML) and AI for the years 2016 to 2022 in the English language. A total of 210 articles were retrieved that were screened with PRISMA for abstract, year, language, context, and proximal relevance to research aims. A thematic analysis was carried out to extract findings from the included studies.

RESULTS

Twenty articles were selected for data abstraction and analysis, with three consistent themes emerging from these articles. In particular, eleven articles address the current state-of-the-art application of AI in diagnosing and predicting health conditions such as behavioral and mental health, cancer, syndromic and metabolic diseases. Five articles highlight the specific challenges of AI deployment in pediatric medicines: data security, handling, authentication, and validation. Four articles set out future opportunities for AI to be adapted: the incorporation of Big Data, cloud computing, precision medicine, and clinical decision support systems. These studies collectively critically evaluate the potential of AI in overcoming current barriers to adoption.

CONCLUSION

AI is proving disruptive within pediatric medicine and is presently associated with challenges, opportunities, and the need for explainability. AI should be viewed as a tool to enhance and support clinical decision-making rather than a substitute for human judgement and expertise. Future research should consequently focus on obtaining comprehensive data to ensure the generalizability of research findings.

Early prediction of mental health problems following military deployment: Integrating pre- and post-deployment factors in neural network models

Military personnel deployed to war zones are at increased risk of mental health problems such as posttraumatic stress disorder (PTSD) or depression. Early pre- or post-deployment identification of those at highest risk of such problems is crucial to target intervention to those in need. However, sufficiently accurate models predicting objectively assessed mental health outcomes have not been put forward. In a sample consisting of all Danish military personnel who deployed to war zones for the first (N = 27,594), second (N = 11,083) and third (N = 5,161) time between 1992 and 2013, we apply neural networks to predict psychiatric diagnoses or use of psychotropic medicine in the years following deployment. Models are based on pre-deployment registry data alone or on pre-deployment registry data in combination with post-deployment questionnaire data on deployment experiences or early post-deployment reactions. Further, we identified the most central predictors of importance for the first, second, and third deployment. Models based on pre-deployment registry data alone had lower accuracy (AUCs ranging from 0.61 (third deployment) to 0.67 (first deployment)) than models including pre- and post-deployment data (AUCs ranging from 0.70 (third deployment) to 0.74 (first deployment)). Age at deployment, deployment year and previous physical trauma were important across deployments. Post-deployment predictors varied across deployments but included deployment exposures as well as early post-deployment symptoms. The results suggest that neural network models combining pre- and early post-deployment data can be utilized for screening tools that identify individuals at risk of severe mental health problems in the years following military deployment.

A computational model for learning from repeated traumatic experiences under uncertainty

Traumatic events can lead to lifelong, inflexible adaptations in threat perception and behavior, which characterize posttraumatic stress disorder (PTSD). This process involves associations between sensory cues and internal states of threat and then generalization of the threat responses to previously neutral cues. However, most formulations neglect adaptations to threat that are not specific to those associations. To incorporate nonassociative responses to threat, we propose a computational theory of PTSD based on adaptation to the frequency of traumatic events by using a reinforcement learning momentum model. Recent threat prediction errors generate momentum that influences subsequent threat perception in novel contexts. This model fits primary data acquired from a mouse model of PTSD, in which unpredictable footshocks in one context accelerate threat learning in a novel context. The theory is consistent with epidemiological data that show that PTSD incidence increases with the number of traumatic events, as well as the disproportionate impact of early life trauma. Because the theory proposes that PTSD relates to the average of recent threat prediction errors rather than the strength of a specific association, it makes novel predictions for the treatment of PTSD.

Applications of artificial intelligence-machine learning for detection of stress: a critical overview

Psychological distress is a major contributor to human physiology and pathophysiology, and it has been linked to several conditions, such as auto-immune diseases, metabolic syndrome, sleep disorders, and suicidal thoughts and inclination. Therefore, early detection and management of chronic stress is crucial for the prevention of several diseases. Artificial intelligence (AI) and Machine Learning (ML) have promoted a paradigm shift in several areas of biomedicine including diagnosis, monitoring, and prognosis of disease. Here, our review aims to present some of the AI and ML applications for solving biomedical issues related to psychological stress. We provide several lines of evidence from previous studies highlighting that AI and ML have been able to predict stress and detect the brain normal states vs. abnormal states (notably, in post-traumatic stress disorder (PTSD)) with accuracy around 90%. Of note, AI/ML-driven technology applied to identify ubiquitously present stress exposure may not reach its full potential, unless future analytics focus on detecting prolonged distress through such technology instead of merely assessing stress exposure. Moving forward, we propose that a new subcategory of AI methods called Swarm Intelligence (SI) can be used towards detecting stress and PTSD. SI involves ensemble learning techniques to efficiently solve a complex problem, such as stress detection, and it offers particular strength in clinical settings, such as privacy preservation. We posit that AI and ML approaches will be beneficial for the medical and patient community when applied to predict and assess stress levels. Last, we encourage additional research to bring AI and ML into the standard clinical practice for diagnostics in the not-too-distant future.

Derivation and Validation of a Brief Emergency Department-Based Prediction Tool for Posttraumatic Stress After Motor Vehicle Collision

STUDY OBJECTIVE

To derive and initially validate a brief bedside clinical decision support tool that identifies emergency department (ED) patients at high risk of substantial, persistent posttraumatic stress symptoms after a motor vehicle collision.

METHODS

Derivation (n=1,282, 19 ED sites) and validation (n=282, 11 separate ED sites) data were obtained from adults prospectively enrolled in the Advancing Understanding of RecOvery afteR traumA study who were discharged from the ED after motor vehicle collision-related trauma. The primary outcome was substantial posttraumatic stress symptoms at 3 months (Posttraumatic Stress Disorder Checklist for Diagnostic and Statistical Manual of Mental Disorders-5 ≥38). Logistic regression derivation models were evaluated for discriminative ability using the area under the curve and the accuracy of predicted risk probabilities (Brier score). Candidate posttraumatic stress predictors assessed in these models (n=265) spanned a range of sociodemographic, baseline health, peritraumatic, and mechanistic domains. The final model selection was based on performance and ease of administration.

RESULTS

Significant 3-month posttraumatic stress symptoms were common in the derivation (27%) and validation (26%) cohort. The area under the curve and Brier score of the final 8-question tool were 0.82 and 0.14 in the derivation cohort and 0.76 and 0.17 in the validation cohort.

CONCLUSION

This simple 8-question tool demonstrates promise to risk-stratify individuals with substantial posttraumatic stress symptoms who are discharged to home after a motor vehicle collision. Both external validation of this instrument, and work to further develop more accurate tools, are needed. Such tools might benefit public health by enabling the conduct of preventive intervention trials and assisting the growing number of EDs that provide services to trauma survivors aimed at promoting psychological recovery.

Superstorm Sandy exposure in utero is associated with neurobehavioral phenotypes and brain structure alterations in childhood: A machine learning approach

Introduction

Prenatal maternal stress (PNMS), including exposure to natural disasters, has been shown to serve as a risk factor for future child psychopathology and suboptimal brain development, particularly among brain regions shown to be sensitive to stress and trauma exposure. However, statistical approaches deployed in most studies are usually constrained by a limited number of variables for the sake of statistical power. Explainable machine learning, on the other hand, enables the study of high data dimension and offers novel insights into the prominent subset of behavioral phenotypes and brain regions most susceptible to PNMS. In the present study, we aimed to identify the most important child neurobehavioral and brain features associated with in utero exposure to Superstorm Sandy (SS).

Methods

By leveraging an explainable machine learning technique, the Shapley additive explanations method, we tested the marginal feature effect on SS exposures and examined the individual variable effects on disaster exposure.

Results

Results show that certain brain regions are especially sensitive to in utero exposure to SS. Specifically, in utero SS exposure was associated with larger gray matter volume (GMV) in the right caudate, right hippocampus, and left amygdala and smaller GMV in the right parahippocampal gyrus. Additionally, higher aggression scores at age 5 distinctly correlated with SS exposure.

Discussion

These findings suggest in utero SS exposure may be associated with greater aggression and suboptimal developmental alterations among various limbic and basal ganglia brain regions.

Acute stress and PTSD among trauma-exposed children and adolescents: Computational prediction and interpretation

BACKGROUND

Youth receiving medical care for injury are at risk of PTSD. Therefore, accurate prediction of chronic PTSD at an early stage is needed. Machine learning (ML) offers a promising approach to precise prediction and interpretation.

AIMS

The study proposes a clinically useful predictive model for PTSD 6-12 months after injury, analyzing the relationship among predictors, and between predictors and outcomes.

METHODS

A ML approach was utilized to train models based on 1167 children and adolescents of nine perspective studies. Demographics, trauma characteristics and acute traumatic stress (ASD) symptoms were used as initial predictors. PTSD diagnosis at six months was derived using DSM-IV PTSD diagnostic criteria. Models were validated on external datasets. Shapley value and partial dependency plot (PDP) were applied to interpret the final model.

RESULTS

A random forest model with 13 predictors (age, ethnicity, trauma type, intrusive memories, nightmares, reliving, distress, dissociation, cognitive avoidance, sleep, irritability, hypervigilance and startle) yielded F-scores of.973,0.902 and.961 with training and two external datasets. Shapley values were calculated for individual and grouped predictors. A cumulative effect for intrusion symptoms was observed. PDP showed a non-linear relationship between age and PTSD, and between ASD symptom severity and PTSD. A 43 % difference in the risk between non-minority and minority ethnic groups was detected.

CONCLUSIONS

A ML model demonstrated excellent classification performance and good potential for clinical utility, using a few easily obtainable variables. Model interpretation gave a comprehensive quantitative analysis on the operations among predictors, in particular ASD symptoms.

Who will respond to intensive PTSD treatment? A machine learning approach to predicting response prior to starting treatment

Despite the established effectiveness of evidence-based PTSD treatments, not everyone responds the same. Specifically, some individuals respond early while others respond minimally throughout treatment. Our ability to predict these trajectories at baseline has been limited. Predicting which individuals will respond to a certain type of treatment can significantly reduce short- and long-term costs and increase the ability to preemptively match individuals with treatments to which they are most likely to respond. In the present study, we examined whether veterans' responses to a 3-week Cognitive Processing Therapy-based intensive PTSD treatment program could be accurately predicted prior to the first session. Using a sample of 432 veterans, and a wide range of demographic and clinical data collected during intake, we assessed six machine learning and statistical methods and their ability to predict fast and minimal responders prior to treatment initiation. For fast response classification, gradient boosted models (GBM) had the highest AUC-PR (0.466). For minimal response classification, elastic net (EN) had the highest mean CV AUC-PR (0.628). Using the best performing classifiers, we were able to predict both fast and minimal responders prior to starting treatment with relatively high AUC-ROC of 0.765 (GBM) and 0.826 (EN), respectively. These results may inform treatment modifications, although the accuracy may not be sufficient for clinicians to base inclusion/exclusion decisions entirely on the classifiers. Future research should evaluate whether these classifiers can be expanded to predict to which treatment type(s) an individual is most likely to respond based on various clinical, circumstantial, and biological features.

Mental health progress requires causal diagnostic nosology and scalable causal discovery

Nine hundred and seventy million individuals across the globe are estimated to carry the burden of a mental disorder. Limited progress has been achieved in alleviating this burden over decades of effort, compared to progress achieved for many other medical disorders. Progress on outcome improvement for all medical disorders, including mental disorders, requires research capable of discovering causality at sufficient scale and speed, and a diagnostic nosology capable of encoding the causal knowledge that is discovered. Accordingly, the field's guiding paradigm limits progress by maintaining: (a) a diagnostic nosology (DSM-5) with a profound lack of causality; (b) a misalignment between mental health etiologic research and nosology; (c) an over-reliance on clinical trials beyond their capabilities; and (d) a limited adoption of newer methods capable of discovering the complex etiology of mental disorders. We detail feasible directions forward, to achieve greater levels of progress on improving outcomes for mental disorders, by: (a) the discovery of knowledge on the complex etiology of mental disorders with application of Causal Data Science methods; and (b) the encoding of the etiological knowledge that is discovered within a causal diagnostic system for mental disorders.

A systematic literature review of AI-based digital decision support systems for post-traumatic stress disorder

OBJECTIVE

Over the last decade, an increase in research on medical decision support systems has been observed. However, compared to other disciplines, decision support systems in mental health are still in the minority, especially for rare diseases like post-traumatic stress disorder (PTSD). We aim to provide a comprehensive analysis of state-of-the-art digital decision support systems (DDSSs) for PTSD.

METHODS

Based on our systematic literature review of DDSSs for PTSD, we created an analytical framework using thematic analysis for feature extraction and quantitative analysis for the literature. Based on this framework, we extracted information around the medical domain of DDSSs, the data used, the technology used for data collection, user interaction, decision-making, user groups, validation, decision type and maturity level. Extracting data for all of these framework dimensions ensures consistency in our analysis and gives a holistic overview of DDSSs.

RESULTS

Research on DDSSs for PTSD is rare and primarily deals with the algorithmic part of DDSSs (n = 17). Only one DDSS was found to be a usable product. From a data perspective, mostly checklists or questionnaires were used (n = 9). While the median sample size of 151 was rather low, the average accuracy was 82%. Validation, excluding algorithmic accuracy (like user acceptance), was mostly neglected, as was an analysis concerning possible user groups.

CONCLUSION

Based on a systematic literature review, we developed a framework covering all parts (medical domain, data used, technology used for data collection, user interaction, decision-making, user groups, validation, decision type and maturity level) of DDSSs. Our framework was then used to analyze DDSSs for post-traumatic stress disorder. We found that DDSSs are not ready-to-use products but are mostly algorithms based on secondary datasets. This shows that there is still a gap between technical possibilities and real-world clinical work.

An Explainable Intelligence Driven Query Prioritization Using Balanced Decision Tree Approach for Multi-Level Psychological Disorders Assessment

Human emotions affect psychological health to a great level. Positive emotions relate to health improvement; whereas negative emotions may aggravate psychological disorders such as anxiety, stress, and depression. Although there exist several computational methods to predict psychological disorders, most of them provide a black-box view of uncertainty. This research involves developing a novel predictive model for multi class psychological risk recognition with an accurate explainable interface. Standard questionnaires are utilized as data set and a new approach called a Q-Prioritization is employed to drop insignificant questions from the data set. Moreover, a novel balanced decision tree method based on repetitive oversampling is applied for the training and testing of the model. Predictive nature along with its contributing factors are interpreted with three techniques such as permuted feature importance, contrastive explanation, and counterfactual method, which together form a reasoning engine. The prediction outcome generated an impressive performance with an aggregated accuracy of 98.25%. The mean precision, recall, and F-score metric recorded were 0.98, 0.977, and 0.979, respectively. Also, it was noted that without applying Q-Prioritization, the accuracy significantly drops to 90.25%. The error rate observed with our model was only 0.026. The proposed multi-level psychological disorder predictive model can successfully serve as an assistive deployment for medical experts in the effective treatment of mental health.

Diagnosing post-traumatic stress disorder using electronic medical record data

This study proposes a predictive model that uses structured data and unstructured narrative notes from Electronic Medical Records to accurately identify patients diagnosed with Post-Traumatic Stress Disorder (PTSD). We utilize data from primary care clinicians participating in the Manitoba Primary Care Research Network (MaPCReN) representing 154,118 patients. A reference sample of 195 patients that had their PTSD diagnosis confirmed using a manual chart review of structured data and narrative notes, and PTSD negative patients is used as the gold standard data for model training, validation and testing. We assess structured and unstructured data from eight tables in the MaPCReN namely, patient demographics, disease case, examinations, medication, billing records, health condition, risk factors, and encounter notes. Feature engineering is applied to convert data into proper representation for predictive modeling. We explore serial and parallel mixed data models that are trained on both structured and unstructured data to identify PTSD. Model performances were calculated based on a highly skewed hold-out test dataset. The serial model that uses both structured and text data as input, yielded the highest values in sensitivity (0.77), F-measure (0.76), and AUC (0.88) and the parallel model that uses both structured and text data as the input obtained the highest positive predicted value (PPV) (0.75). Diseases such as PTSD are difficult to diagnose. Information recorded in the chart note over multiple visits of the patients with the primary care physicians has higher predictive power than structured data and combining these two data types can increase the predictive capabilities of machine learning models in diagnosing PTSD. While the deep-learning model outperformed the traditional ensemble model in processing text data, the ensemble classifier obtained better results in ingesting a combination of features obtained from both data types in the serial mixed model. The study demonstrated that unstructured encounter notes enhance a model's ability to identify patients diagnosed with PTSD. These findings can enhance quality improvement, research, and disease surveillance related to PTSD in primary care populations.

Development and Validation of a Model to Predict Posttraumatic Stress Disorder and Major Depression After a Motor Vehicle Collision

IMPORTANCE

A substantial proportion of the 40 million people in the US who present to emergency departments (EDs) each year after traumatic events develop posttraumatic stress disorder (PTSD) or major depressive episode (MDE). Accurately identifying patients at high risk in the ED would facilitate the targeting of preventive interventions.

OBJECTIVES

To develop and validate a prediction tool based on ED reports after a motor vehicle collision to predict PTSD or MDE 3 months later.

DESIGN, SETTING, AND PARTICIPANTS

The Advancing Understanding of Recovery After Trauma (AURORA) study is a longitudinal study that examined adverse posttraumatic neuropsychiatric sequalae among patients who presented to 28 US urban EDs in the immediate aftermath of a traumatic experience. Enrollment began on September 25, 2017. The 1003 patients considered in this diagnostic/prognostic report completed 3-month assessments by January 31, 2020. Each patient received a baseline ED assessment along with follow-up self-report surveys 2 weeks, 8 weeks, and 3 months later. An ensemble machine learning method was used to predict 3-month PTSD or MDE from baseline information. Data analysis was performed from November 1, 2020, to May 31, 2021.

MAIN OUTCOMES AND MEASURES

The PTSD Checklist for DSM-5 was used to assess PTSD and the Patient Reported Outcomes Measurement Information System Depression Short-Form 8b to assess MDE.

RESULTS

A total of 1003 patients (median [interquartile range] age, 34.5 [24-43] years; 715 [weighted 67.9%] female; 100 [weighted 10.7%] Hispanic, 537 [weighted 52.7%] non-Hispanic Black, 324 [weighted 32.2%] non-Hispanic White, and 42 [weighted 4.4%] of non-Hispanic other race or ethnicity were included in this study. A total of 274 patients (weighted 26.6%) met criteria for 3-month PTSD or MDE. An ensemble machine learning model restricted to 30 predictors estimated in a training sample (patients from the Northeast or Midwest) had good prediction accuracy (mean [SE] area under the curve [AUC], 0.815 [0.031]) and calibration (mean [SE] integrated calibration index, 0.040 [0.002]; mean [SE] expected calibration error, 0.039 [0.002]) in an independent test sample (patients from the South). Patients in the top 30% of predicted risk accounted for 65% of all 3-month PTSD or MDE, with a mean (SE) positive predictive value of 58.2% (6.4%) among these patients at high risk. The model had good consistency across regions of the country in terms of both AUC (mean [SE], 0.789 [0.025] using the Northeast as the test sample and 0.809 [0.023] using the Midwest as the test sample) and calibration (mean [SE] integrated calibration index, 0.048 [0.003] using the Northeast as the test sample and 0.024 [0.001] using the Midwest as the test sample; mean [SE] expected calibration error, 0.034 [0.003] using the Northeast as the test sample and 0.025 [0.001] using the Midwest as the test sample). The most important predictors in terms of Shapley Additive Explanations values were symptoms of anxiety sensitivity and depressive disposition, psychological distress in the 30 days before motor vehicle collision, and peritraumatic psychosomatic symptoms.

CONCLUSIONS AND RELEVANCE

The results of this study suggest that a short set of questions feasible to administer in an ED can predict 3-month PTSD or MDE with good AUC, calibration, and geographic consistency. Patients at high risk can be identified in the ED for targeting if cost-effective preventive interventions are developed.

Prediction of anxiety disorders using a feature ensemble based bayesian neural network

Anxiety disorders are common among youth, posing risks to physical and mental health development. Early screening can help identify such disorders and pave the way for preventative treatment. To this end, the Youth Online Diagnostic Assessment (YODA) tool was developed and deployed to predict youth disorders using online screening questionnaires filled by parents. YODA facilitated collection of several novel unique datasets of self-reported anxiety disorder symptoms. Since the data is self-reported and often noisy, feature selection needs to be performed on the raw data to improve accuracy. However, a single set of selected features may not be informative enough. Consequently, in this work we propose and evaluate a novel feature ensemble based Bayesian Neural Network (FE-BNN) that exploits an ensemble of features for improving the accuracy of disorder predictions. We evaluate the performance of FE-BNN on three disorder-specific datasets collected by YODA. Our method achieved the AUC of 0.8683, 0.8769, 0.9091 for the predictions of Separation Anxiety Disorder, Generalized Anxiety Disorder and Social Anxiety Disorder, respectively. These results provide initial evidence that our method outperforms the original diagnostic scoring function of YODA and several other baseline methods for three anxiety disorders, which can practically help prioritizing diagnostic interviews. Our promising results call for investigation of interpretable methods maintaining high predictive accuracy.

Can the adverse childhood experiences (ACEs) checklist be utilized to predict emergency department visits among children and adolescents?

Background

Extensive literature has shown an association of Adverse Childhood Experiences (ACEs) with adverse health outcomes; however, its ability to predict events or stratify risks is less known. Individuals with mental illness and ACE exposure have been shown to visit emergency departments (ED) more often than those in the general population. This study thus examined the ability of the ACEs checklist to predict ED visits within the subsequent year among children and adolescents presenting to mental health clinics with pre-existing mental health issues.

Methods

The study analyzed linked data (n = 6100) from two databases provided by Alberta Health Services (AHS). The Regional Access and Intake System (RAIS 2016–2018) database provided data on the predictors (ACE items, age, sex, residence, mental health program type, and primary diagnosis) regarding children and adolescents (aged 0–17 years) accessing addiction and mental health services within Calgary Zone, and the National Ambulatory Care Reporting System (NACRS 2016–2019) database provided data on ED visits. A 25% random sample of the data was reserved for validation purposes. Two Least Absolute Shrinkage and Selection Operator (LASSO) logistic regression models, each employing a different method to tune the shrinkage parameter lambda (namely cross-validated and adaptive) and performing 10-fold cross-validation for a set of 100 lambdas in each model were examined.

Results

The adaptive LASSO model had a slightly better fit in the validation dataset than the cross-validated model; however, it still demonstrated poor discrimination (AUC 0.60, sensitivity 37.8%, PPV 49.6%) and poor calibration (over-triaged in low-risk and under-triaged in high-risk subgroups). The model’s poor performance was evident from an out-of-sample deviance ratio of − 0.044.

Conclusion

The ACEs checklist did not perform well in predicting ED visits among children and adolescents with existing mental health concerns. The diverse causes of ED visits may have hindered accurate predictions, requiring more advanced statistical procedures. Future studies exploring other machine learning approaches and including a more extensive set of childhood adversities and other important predictors may produce better predictions. Furthermore, despite highly significant associations being observed, ACEs may not be deterministic in predicting health-related events at the individual level, such as general ED use.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12874-021-01392-w.

Deep learning model of fMRI connectivity predicts PTSD symptom trajectories in recent trauma survivors

Early intervention following exposure to a traumatic life event could change the clinical path from the development of post traumatic stress disorder (PTSD) to recovery, hence the interest in early detection and underlying biological mechanisms involved in the development of post traumatic sequelae. We introduce a novel end-to-end neural network that employs resting-state and task-based functional MRI (fMRI) datasets, obtained one month after trauma exposure, to predict PTSD symptoms at one-, six- and fourteen-months after the exposure. FMRI data, as well as PTSD status and symptoms, were collected from adults at risk for PTSD development, after admission to emergency room following a traumatic event. Our computational method utilized a per-region encoder to extract brain regions embedding, which were subsequently updated by applying the algorithmic technique of pairwise attention. The affinities obtained between each pair of regions were combined to create a pairwise co-activation map used to perform multi-label classification. The results demonstrate that the novel method's performance in predicting PTSD symptoms, in a prospective manner, outperforms previous analytical techniques reported in the fMRI literature, all trained on the same dataset. We further show a high predictive ability for predicting PTSD symptom clusters and PTSD persistence. To the best of our knowledge, this is the first deep learning method applied on fMRI data with respect to prospective clinical outcomes, to predict PTSD status, severity and symptom clusters. Future work could further delineate the mechanisms that underlie such a prediction, and potentially improve single patient characterization.

Improving Mental Health Services: A 50-Year Journey from Randomized Experiments to Artificial Intelligence and Precision Mental Health

This conceptual paper describes the current state of mental health services, identifies critical problems, and suggests how to solve them. I focus on the potential contributions of artificial intelligence and precision mental health to improving mental health services. Toward that end, I draw upon my own research, which has changed over the last half century, to highlight the need to transform the way we conduct mental health services research. I identify exemplars from the emerging literature on artificial intelligence and precision approaches to treatment in which there is an attempt to personalize or fit the treatment to the client in order to produce more effective interventions.

Prospective prediction of PTSD diagnosis in a nationally representative sample using machine learning

BACKGROUND

Recent research has identified a number of pre-traumatic, peri-traumatic and post-traumatic psychological and ecological factors that put an individual at increased risk for developing PTSD following a life-threatening event. While these factors have been found to be associated with PTSD in univariate analyses, the complex interactions of these risk factors and how they contribute to individual trajectories of the illness are not yet well understood. In this study, we examine the impact of prior trauma, psychopathology, sociodemographic characteristics, community and environmental information, on PTSD onset in a nationally representative sample of adults in the United States, using machine learning methods to establish the relative contributions of each variable.

METHODS

Individual risk factors identified in Waves 1 of the National Epidemiologic Survey on Alcohol and Related Conditions (NESARC) were combined with community-level data for the years concurrent to the NESARC Wave 1 (n = 43,093) and 2 (n = 34,653) surveys. Machine learning feature selection and classification analyses were used at the national level to create models using individual- and community-level variables that would best predict the new onset of PTSD at Wave 2.

RESULTS

Our classification algorithms yielded 89.7 to 95.6% accuracy for predicting new onset of PTSD at Wave 2. A prior diagnosis of DSM-IV-TR Borderline Personality Disorder, Major Depressive Disorder or Anxiety Disorder conferred the greatest relative influence in new diagnosis of PTSD. Distal risk factors such as prior psychiatric diagnosis accounted for significantly greater relative risk than proximal factors (such as adverse event exposure).

CONCLUSIONS

Our findings show that a machine learning classification approach can successfully integrate large numbers of known risk factors for PTSD into stronger models that account for high-dimensional interactions and collinearity between variables. We discuss the implications of these findings as pertaining to the targeted mobilization emergency mental health resources. These findings also inform the creation of a more comprehensive risk assessment profile to the likelihood of developing PTSD following an extremely adverse event.

A Machine-Based Prediction Model of ADHD Using CPT Data

Despite the popularity of the continuous performance test (CPT) in the diagnosis of attention-deficit/hyperactivity disorder (ADHD), its specificity, sensitivity, and ecological validity are still debated. To address some of the known shortcomings of traditional analysis and interpretation of CPT data, the present study applied a machine learning-based model (ML) using CPT indices for the Prediction of ADHD.Using a retrospective factorial fitting, followed by a bootstrap technique, we trained, cross-validated, and tested learning models on CPT performance data of 458 children aged 6–12 years (213 children with ADHD and 245 typically developed children). We used the MOXO-CPT version that included visual and auditory stimuli distractors. Results showed that the ML proposed model performed better and had a higher accuracy than the benchmark approach that used clinical data only. Using the CPT total score (that included all four indices: Attention, Timeliness, Hyperactivity, and Impulsiveness), as well as four control variables [age, gender, day of the week (DoW), time of day (ToD)], provided the most salient information for discriminating children with ADHD from their typically developed peers. This model had an accuracy rate of 87%, a sensitivity rate of 89%, and a specificity rate of 84%. This performance was 34% higher than the best-achieved accuracy of the benchmark model. The ML detection model could classify children with ADHD with high accuracy based on CPT performance. ML model of ADHD holds the promise of enhancing, perhaps complementing, behavioral assessment and may be used as a supportive measure in the evaluation of ADHD.

Computational causal discovery for post-traumatic stress in police officers

This article reports on a study aimed to elucidate the complex etiology of post-traumatic stress (PTS) in a longitudinal cohort of police officers, by applying rigorous computational causal discovery (CCD) methods with observational data. An existing observational data set was used, which comprised a sample of 207 police officers who were recruited upon entry to police academy training. Participants were evaluated on a comprehensive set of clinical, self-report, genetic, neuroendocrine and physiological measures at baseline during academy training and then were re-evaluated at 12 months after training was completed. A data-processing pipeline-the Protocol for Computational Causal Discovery in Psychiatry (PCCDP)-was applied to this data set to determine a causal model for PTS severity. A causal model of 146 variables and 345 bivariate relations was discovered. This model revealed 5 direct causes and 83 causal pathways (of four steps or less) to PTS at 12 months of police service. Direct causes included single-nucleotide polymorphisms (SNPs) for the Histidine Decarboxylase (HDC) and Mineralocorticoid Receptor (MR) genes, acoustic startle in the context of low perceived threat during training, peritraumatic distress to incident exposure during first year of service, and general symptom severity during training at 1 year of service. The application of CCD methods can determine variables and pathways related to the complex etiology of PTS in a cohort of police officers. This knowledge may inform new approaches to treatment and prevention of critical incident related PTS.

Applicability of an Automated Model and Parameter Selection in the Prediction of Screening-Level PTSD in Danish Soldiers Following Deployment: Development Study of Transferable Predictive Models Using Automated Machine Learning

BACKGROUND

Posttraumatic stress disorder (PTSD) is a relatively common consequence of deployment to war zones. Early postdeployment screening with the aim of identifying those at risk for PTSD in the years following deployment will help deliver interventions to those in need but have so far proved unsuccessful.

OBJECTIVE

This study aimed to test the applicability of automated model selection and the ability of automated machine learning prediction models to transfer across cohorts and predict screening-level PTSD 2.5 years and 6.5 years after deployment.

METHODS

Automated machine learning was applied to data routinely collected 6-8 months after return from deployment from 3 different cohorts of Danish soldiers deployed to Afghanistan in 2009 (cohort 1, N=287 or N=261 depending on the timing of the outcome assessment), 2010 (cohort 2, N=352), and 2013 (cohort 3, N=232).

RESULTS

Models transferred well between cohorts. For screening-level PTSD 2.5 and 6.5 years after deployment, random forest models provided the highest accuracy as measured by area under the receiver operating characteristic curve (AUC): 2.5 years, AUC=0.77, 95% CI 0.71-0.83; 6.5 years, AUC=0.78, 95% CI 0.73-0.83. Linear models performed equally well. Military rank, hyperarousal symptoms, and total level of PTSD symptoms were highly predictive.

CONCLUSIONS

Automated machine learning provided validated models that can be readily implemented in future deployment cohorts in the Danish Defense with the aim of targeting postdeployment support interventions to those at highest risk for developing PTSD, provided the cohorts are deployed on similar missions.

Classifying heterogeneous presentations of PTSD via the default mode, central executive, and salience networks with machine learning

Intrinsic connectivity networks (ICNs), including the default mode network (DMN), the central executive network (CEN), and the salience network (SN) have been shown to be aberrant in patients with posttraumatic stress disorder (PTSD). The purpose of the current study was to a) compare ICN functional connectivity between PTSD, dissociative subtype PTSD (PTSD+DS) and healthy individuals; and b) to examine the use of multivariate machine learning algorithms in classifying PTSD, PTSD+DS, and healthy individuals based on ICN functional activation. Our neuroimaging dataset consisted of resting-state fMRI scans from 186 participants [PTSD (n = 81); PTSD + DS (n = 49); and healthy controls (n = 56)]. We performed group-level independent component analyses to evaluate functional connectivity differences within each ICN. Multiclass Gaussian Process Classification algorithms within PRoNTo software were then used to predict the diagnosis of PTSD, PTSD+DS, and healthy individuals based on ICN functional activation. When comparing the functional connectivity of ICNs between PTSD, PTSD+DS and healthy controls, we found differential patterns of connectivity to brain regions involved in emotion regulation, in addition to limbic structures and areas involved in self-referential processing, interoception, bodily self-consciousness, and depersonalization/derealization. Machine learning algorithms were able to predict with high accuracy the classification of PTSD, PTSD+DS, and healthy individuals based on ICN functional activation. Our results suggest that alterations within intrinsic connectivity networks may underlie unique psychopathology and symptom presentation among PTSD subtypes. Furthermore, the current findings substantiate the use of machine learning algorithms for classifying subtypes of PTSD illness based on ICNs.

Identifying predictors of probable posttraumatic stress disorder in children and adolescents with earthquake exposure: A longitudinal study using a machine learning approach

BACKGROUND

Evidence has identified risk factors associated with individuals with trauma exposure who develop posttraumatic stress disorder (PTSD). How to combine risk factors to predict probable PTSD in young survivors using machine learning is limited. The study aimed to integrated multiple measures at 2 weeks after the earthquake using machine learning for the prediction of probable PTSD at 3 months after earthquake.

METHODS

A total of 2099 young survivors with earthquake exposure were included. We integrated multiple domains of variables to 'train' a machine learning algorithm (XGBoost). Thirty-one combination types were implemented and evaluated. The resulting XGBoost was utilized in identifying individual participants as either probable PTSD or no PTSD.

RESULTS

Any combination type predicted young survivor probable PTSD, with prediction accuracies ranging between 66%-80% (p < 0.05). In particular, the combination of earthquake experience, everyday functioning, somatic symptoms and sleeping correctly predicted 683 out of 802 cases of probable PTSD, translating to a classical accuracy of 74.476% (85.156% sensitivity and 60.366% specificity) and an area under the curve of 0.80. The most relevant variables (e.g. age, sex, property loss and a sedentary lifestyle) revealed in the present study.

LIMITATIONS

Participants from a specific district might limit the generalizability of our results. Self-report questionnaires and non-standardized measures were used to assess symptoms.

CONCLUSION

Detection of probable PTSD according to self-reported measurement data is feasible, may improve operational efficiencies via enabling targeted intervention, before manifestation of symptoms.

The use of machine learning techniques in trauma-related disorders: a systematic review

Establishing the diagnosis of trauma-related disorders such as Acute Stress Disorder (ASD) and Posttraumatic Stress Disorder (PTSD) have always been a challenge in clinical practice and in academic research, due to clinical and biological heterogeneity. Machine learning (ML) techniques can be applied to improve classification of disorders, to predict outcomes or to determine person-specific treatment selection. We aim to review the existing literature on the use of machine learning techniques in the assessment of subjects with ASD or PTSD. We systematically searched PubMed, Embase and Web of Science for articles published in any language up to May 2019. We found 806 abstracts and included 49 studies in our review. Most of the included studies used multiple levels of biological data to predict risk factors or to identify early symptoms related to PTSD. Other studies used ML classification techniques to distinguish individuals with ASD or PTSD from other psychiatric disorder or from trauma-exposed and healthy controls. We also found studies that attempted to define outcome profiles using clustering techniques and studies that assessed the relationship among symptoms using network analysis. Finally, we proposed a quality assessment in this review, evaluating methodological and technical features on machine learning studies. We concluded that etiologic and clinical heterogeneity of ASD/PTSD patients is suitable to machine learning techniques and a major challenge for the future is to use it in clinical practice for the benefit of patients in an individual level.

Machine learning multivariate pattern analysis predicts classification of posttraumatic stress disorder and its dissociative subtype: a multimodal neuroimaging approach

BACKGROUND

The field of psychiatry would benefit significantly from developing objective biomarkers that could facilitate the early identification of heterogeneous subtypes of illness. Critically, although machine learning pattern recognition methods have been applied recently to predict many psychiatric disorders, these techniques have not been utilized to predict subtypes of posttraumatic stress disorder (PTSD), including the dissociative subtype of PTSD (PTSD + DS).

METHODS

Using Multiclass Gaussian Process Classification within PRoNTo, we examined the classification accuracy of: (i) the mean amplitude of low-frequency fluctuations (mALFF; reflecting spontaneous neural activity during rest); and (ii) seed-based amygdala complex functional connectivity within 181 participants [PTSD (n = 81); PTSD + DS (n = 49); and age-matched healthy trauma-unexposed controls (n = 51)]. We also computed mass-univariate analyses in order to observe regional group differences [false-discovery-rate (FDR)-cluster corrected p < 0.05, k = 20].

RESULTS

We found that extracted features could predict accurately the classification of PTSD, PTSD + DS, and healthy controls, using both resting-state mALFF (91.63% balanced accuracy, p < 0.001) and amygdala complex connectivity maps (85.00% balanced accuracy, p < 0.001). These results were replicated using independent machine learning algorithms/cross-validation procedures. Moreover, areas weighted as being most important for group classification also displayed significant group differences at the univariate level. Here, whereas the PTSD + DS group displayed increased activation within emotion regulation regions, the PTSD group showed increased activation within the amygdala, globus pallidus, and motor/somatosensory regions.

CONCLUSION

The current study has significant implications for advancing machine learning applications within the field of psychiatry, as well as for developing objective biomarkers indicative of diagnostic heterogeneity.

Machine learning in mental health: a scoping review of methods and applications

BACKGROUND

This paper aims to synthesise the literature on machine learning (ML) and big data applications for mental health, highlighting current research and applications in practice.

METHODS

We employed a scoping review methodology to rapidly map the field of ML in mental health. Eight health and information technology research databases were searched for papers covering this domain. Articles were assessed by two reviewers, and data were extracted on the article's mental health application, ML technique, data type, and study results. Articles were then synthesised via narrative review.

RESULTS

Three hundred papers focusing on the application of ML to mental health were identified. Four main application domains emerged in the literature, including: (i) detection and diagnosis; (ii) prognosis, treatment and support; (iii) public health, and; (iv) research and clinical administration. The most common mental health conditions addressed included depression, schizophrenia, and Alzheimer's disease. ML techniques used included support vector machines, decision trees, neural networks, latent Dirichlet allocation, and clustering.

CONCLUSIONS

Overall, the application of ML to mental health has demonstrated a range of benefits across the areas of diagnosis, treatment and support, research, and clinical administration. With the majority of studies identified focusing on the detection and diagnosis of mental health conditions, it is evident that there is significant room for the application of ML to other areas of psychology and mental health. The challenges of using ML techniques are discussed, as well as opportunities to improve and advance the field.

Machine Learning for Prediction of Posttraumatic Stress and Resilience Following Trauma: An Overview of Basic Concepts and Recent Advances

Posttraumatic stress responses are characterized by a heterogeneity in clinical appearance and etiology. This heterogeneity impacts the field's ability to characterize, predict, and remediate maladaptive responses to trauma. Machine learning (ML) approaches are increasingly utilized to overcome this foundational problem in characterization, prediction, and treatment selection across branches of medicine that have struggled with similar clinical realities of heterogeneity in etiology and outcome, such as oncology. In this article, we review and evaluate ML approaches and applications utilized in the areas of posttraumatic stress, stress pathology, and resilience research, and present didactic information and examples to aid researchers interested in the relevance of ML to their own research. The examined studies exemplify the high potential of ML approaches to build accurate predictive and diagnostic models of posttraumatic stress and stress pathology risk based on diverse sources of available information. The use of ML approaches to integrate high-dimensional data demonstrates substantial gains in risk prediction even when the sources of data are the same as those used in traditional predictive models. This area of research will greatly benefit from collaboration and data sharing among researchers of posttraumatic stress disorder, stress pathology, and resilience.

Applying Machine Learning Techniques to Identify Undiagnosed Patients with Exocrine Pancreatic Insufficiency

BACKGROUND

Exocrine pancreatic insufficiency (EPI) is a serious condition characterized by a lack of functional exocrine pancreatic enzymes and the resultant inability to properly digest nutrients. EPI can be caused by a variety of disorders, including chronic pancreatitis, pancreatic cancer, and celiac disease. EPI remains underdiagnosed because of the nonspecific nature of clinical symptoms, lack of an ideal diagnostic test, and the inability to easily identify affected patients using administrative claims data.

OBJECTIVES

To develop a machine learning model that identifies patients in a commercial medical claims database who likely have EPI but are undiagnosed.

METHODS

A machine learning algorithm was developed in Scikit-learn, a Python module. The study population, selected from the 2014 Truven MarketScan® Commercial Claims Database, consisted of patients with EPI-prone conditions. Patients were labeled with 290 condition category flags and split into actual positive EPI cases, actual negative EPI cases, and unlabeled cases. The study population was then randomly divided into a training subset and a testing subset. The training subset was used to determine the performance metrics of 27 models and to select the highest performing model, and the testing subset was used to evaluate performance of the best machine learning model.

RESULTS

The study population consisted of 2088 actual positive EPI cases, 1077 actual negative EPI cases, and 437 530 unlabeled cases. In the best performing model, the precision, recall, and accuracy were 0.91, 0.80, and 0.86, respectively. The best-performing model estimated that the number of patients likely to have EPI was about 12 times the number of patients directly identified as EPI-positive through a claims analysis in the study population. The most important features in assigning EPI probability were the presence or absence of diagnosis codes related to pancreatic and digestive conditions.

CONCLUSIONS

Machine learning techniques demonstrated high predictive power in identifying patients with EPI and could facilitate an enhanced understanding of its etiology and help to identify patients for possible diagnosis and treatment.

Editorial: In the Causal Labyrinth: Finding the Path From Early Trauma to Neurodevelopment

Exposure to traumatic events early in life has been associated with significant adverse neurodevelopmental consequences.^1,2 Ideally, protecting children from trauma would prevent these consequences. Unfortunately, once a child is exposed to early trauma, the only means of prevention requires implementing effective interventions toward the mechanisms known to have caused the consequences. Therefore, it is imperative to carefully define the true mechanistic pathways from which the neurodevelopmental consequences of early trauma result. The process of determining the correct pathway(s) is not a trivial matter, as is made clear from a valiant effort reported in this issue of the Journal titled, "Early Sexual Trauma Exposure and Neural Response Inhibition in Adolescence and Young Adults: Trajectories of Frontal Theta Oscillations During a Go/No-Go Task."³ This editorial reviews the effort to infer causal mechanism in this article, and introduces a literature that can improve the field's capacity to infer causes from observational data.

Impact of Gender on Child and Adolescent PTSD

PURPOSE OF REVIEW

This review examines the recent literature on biological factors that influence sex differences in posttraumatic stress disorder (PTSD) during childhood and adolescence, focusing on neurobiological, hormonal, and genetic factors that may increase risk in girls.

RECENT FINDINGS

More than 60% of children and adolescents are exposed to traumatic events, and many develop PTSD. There is increasing recognition of gender differences in PTSD, with women having double the rates of the disorder compared to men. These gender differences in symptoms and their underlying neurobiology appear to emerge during adolescence, although it is still unclear which biological mechanisms may play key roles in the development of sex difference. The literature on gender effects in children and adolescents is still in the early stages, and more prospective and longitudinal work is needed; however, estrogen appears to play a key role in increasing risk for PTSD in girls, which emerges in adolescence.