Machine learning improved classification of psychoses using clinical and biological stratification: Update from the bipolar-schizophrenia network for intermediate phenotypes (B-SNIP)

Evidence from comprehensive independent validation studies for smooth pursuit dysfunction as a sensorimotor biomarker for psychosis

Smooth pursuit eye movements are considered a well-established and quantifiable biomarker of sensorimotor function in psychosis research. Identifying psychotic syndromes on an individual level based on neurobiological markers is limited by heterogeneity and requires comprehensive external validation to avoid overestimation of prediction models. Here, we studied quantifiable sensorimotor measures derived from smooth pursuit eye movements in a large sample of psychosis probands (N = 674) and healthy controls (N = 305) using multivariate pattern analysis. Balanced accuracies of 64% for the prediction of psychosis status are in line with recent results from other large heterogenous psychiatric samples. They are confirmed by external validation in independent large samples including probands with (1) psychosis (N = 727) versus healthy controls (N = 292), (2) psychotic (N = 49) and non-psychotic bipolar disorder (N = 36), and (3) non-psychotic affective disorders (N = 119) and psychosis (N = 51) yielding accuracies of 65%, 66% and 58%, respectively, albeit slightly different psychosis syndromes. Our findings make a significant contribution to the identification of biologically defined profiles of heterogeneous psychosis syndromes on an individual level underlining the impact of sensorimotor dysfunction in psychosis.

Supervised machine learning classification of psychosis biotypes based on brain structure: findings from the Bipolar-Schizophrenia network for intermediate phenotypes (B-SNIP)

Traditional diagnostic formulations of psychotic disorders have low correspondence with underlying disease neurobiology. This has led to a growing interest in using brain-based biomarkers to capture biologically-informed psychosis constructs. Building upon our prior work on the B-SNIP Psychosis Biotypes, we aimed to examine whether structural MRI (an independent biomarker not used in the Biotype development) can effectively classify the Biotypes. Whole brain voxel-wise grey matter density (GMD) maps from T1-weighted images were used to train and test (using repeated randomized train/test splits) binary L2-penalized logistic regression models to discriminate psychosis cases (n = 557) from healthy controls (CON, n = 251). A total of six models were evaluated across two psychosis categorization schemes: (i) three Biotypes (B1, B2, B3) and (ii) three DSM diagnoses (schizophrenia (SZ), schizoaffective (SAD) and bipolar (BD) disorders). Above-chance classification accuracies were observed in all Biotype (B1 = 0.70, B2 = 0.65, and B3 = 0.56) and diagnosis (SZ = 0.64, SAD = 0.64, and BD = 0.59) models. However, the only model that showed evidence of specificity was B1, i.e., the model was able to discriminate B1 vs. CON and did not misclassify other psychosis cases (B2 or B3) as B1 at rates above nominal chance. The GMD-based classifier evidence for B1 showed a negative association with an estimate of premorbid general intellectual ability, regardless of group membership, i.e. psychosis or CON. Our findings indicate that, complimentary to clinical diagnoses, the B-SNIP Psychosis Biotypes may offer a promising approach to capture specific aspects of psychosis neurobiology.

Simultaneous clustering and variable selection: A novel algorithm and model selection procedure

The growing availability of high-dimensional data sets offers behavioral scientists an unprecedented opportunity to integrate the information hidden in the novel types of data (e.g., genetic data, social media data, and GPS tracks, etc.,) and thereby obtain a more detailed and comprehensive view towards their research questions. In the context of clustering, analyzing the large volume of variables could potentially result in an accurate estimation or a novel discovery of underlying subgroups. However, a unique challenge is that the high-dimensional data sets likely involve a significant amount of irrelevant variables. These irrelevant variables do not contribute to the separation of clusters and they may mask cluster partitions. The current paper addresses this challenge by introducing a new clustering algorithm, called Cardinality K-means or CKM, and by proposing a novel model selection strategy. CKM is able to perform simultaneous clustering and variable selection with high stability. In two simulation studies and an empirical demonstration with genetic data, CKM consistently outperformed competing methods in terms of recovering cluster partitions and identifying signaling variables. Meanwhile, our novel model selection strategy determines the number of clusters based on a subset of variables that are most likely to be signaling variables. Through a simulation study, this strategy was found to result in a more accurate estimation of the number of clusters compared to the conventional strategy that utilizes the full set of variables. Our proposed CKM algorithm, together with the novel model selection strategy, has been implemented in a freely accessible R package.

The EEG multiverse of schizophrenia

Research on schizophrenia typically focuses on one paradigm for which clear-cut differences between patients and controls are established. Great efforts are made to understand the underlying genetical, neurophysiological, and cognitive mechanisms, which eventually may explain the clinical outcome. One tacit assumption of these "deep rooting" approaches is that paradigms tap into common and representative aspects of the disorder. Here, we analyzed the resting-state electroencephalogram (EEG) of 121 schizophrenia patients and 75 controls. Using multiple signal processing methods, we extracted 194 EEG features. Sixty-nine out of the 194 EEG features showed a significant difference between patients and controls, indicating that these features detect an important aspect of schizophrenia. Surprisingly, the correlations between these features were very low. We discuss several explanations to our results and propose that complementing "deep" with "shallow" rooting approaches might help in understanding the underlying mechanisms of the disorder.

Real-time facial emotion recognition deficits across the psychosis spectrum: A B-SNIP Study

Affective and non-affective psychotic disorders are associated with variable levels of impairment in affective processing, but this domain typically has been examined via presentation of static facial images. We compared performance on a dynamic facial expression identification task across six emotions (sad, fear, surprise, disgust, anger, happy) in individuals with psychotic disorders (bipolar with psychotic features [PBD] = 113, schizoaffective [SAD] = 163, schizophrenia [SZ] = 181) and healthy controls (HC; n = 236) derived from the Bipolar-Schizophrenia Network on Intermediate Phenotypes (B-SNIP). These same individuals with psychotic disorders were also grouped by B-SNIP-derived Biotype (Biotype 1 [B1] = 115, Biotype 2 [B2] = 132, Biotype 3 [B3] = 158), derived from a cluster analysis applied to a large biomarker panel that did not include the current data. Irrespective of the depicted emotion, groups differed in accuracy of emotion identification (P < 0.0001). The SZ group demonstrated lower accuracy versus HC and PBD groups; the SAD group was less accurate than the HC group (Ps < 0.02). Similar overall group differences were evident in speed of identifying emotional expressions. Controlling for general cognitive ability did not eliminate most group differences on accuracy but eliminated almost all group differences on reaction time for emotion identification. Results from the Biotype groups indicated that B1 and B2 had more severe deficits in emotion recognition than HC and B3, meanwhile B3 did not show significant deficits. In sum, this characterization of facial emotion recognition deficits adds to our emerging understanding of social/emotional deficits across the psychosis spectrum.

Use of social media data to diagnose and monitor psychotic disorders: systematic review and perspectives (Preprint)

Background

Schizophrenia is a disease associated with high burden, and improvement in care is necessary. Artificial intelligence (AI) has been used to diagnose several medical conditions as well as psychiatric disorders. However, this technology requires large amounts of data to be efficient. Social media data could be used to improve diagnostic capabilities.

Objective

The objective of our study is to analyze the current capabilities of AI to use social media data as a diagnostic tool for psychotic disorders.

Methods

A systematic review of the literature was conducted using several databases (PubMed, Embase, Cochrane, PsycInfo, and IEEE Xplore) using relevant keywords to search for articles published as of November 12, 2021. We used the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) criteria to identify, select, and critically assess the quality of the relevant studies while minimizing bias. We critically analyzed the methodology of the studies to detect any bias and presented the results.

Results

Among the 93 studies identified, 7 studies were included for analyses. The included studies presented encouraging results. Social media data could be used in several ways to care for patients with schizophrenia, including the monitoring of patients after the first episode of psychosis. We identified several limitations in the included studies, mainly lack of access to clinical diagnostic data, small sample size, and heterogeneity in study quality. We recommend using state-of-the-art natural language processing neural networks, called language models, to model social media activity. Combined with the synthetic minority oversampling technique, language models can tackle the imbalanced data set limitation, which is a necessary constraint to train unbiased classifiers. Furthermore, language models can be easily adapted to the classification task with a procedure called “fine-tuning.”

Conclusions

The use of social media data for the diagnosis of psychotic disorders is promising. However, most of the included studies had significant biases; we therefore could not draw conclusions about accuracy in clinical situations. Future studies need to use more accurate methodologies to obtain unbiased results.

Immune System Abnormalities in Schizophrenia: An Integrative View and Translational Perspectives

The immune system is generally known to be the primary defense mechanism against pathogens. Any pathological conditions are reflected in anomalies in the immune system parameters. Increasing evidence suggests the involvement of immune dysregulation and neuroinflammation in the pathogenesis of schizophrenia. In this systematic review, we summarized the available evidence of abnormalities in the immune system in schizophrenia. We analyzed impairments in all immune system components and assessed the level of bias in the available evidence. It has been shown that schizophrenia is associated with abnormalities in all immune system components: from innate to adaptive immunity and from humoral to cellular immunity. Abnormalities in the immune organs have also been observed in schizophrenia. Evidence of increased C-reactive protein, dysregulation of cytokines and chemokines, elevated levels of neutrophils and autoantibodies, and microbiota dysregulation in schizophrenia have the lowest risk of bias. Peripheral immune abnormalities contribute to neuroinflammation, which is associated with cognitive and neuroanatomical alterations and contributes to the pathogenesis of schizophrenia. However, signs of severe inflammation are observed in only about 1/3 of patients with schizophrenia. Immunological parameters may help identify subgroups of individuals with signs of inflammation who well respond to anti-inflammatory therapy. Our integrative approach also identified gaps in knowledge about immune abnormalities in schizophrenia, and new horizons for the research are proposed.

The Role of Machine Learning in Diagnosing Bipolar Disorder: Scoping Review

Background

Bipolar disorder (BD) is the 10th most common cause of frailty in young individuals and has triggered morbidity and mortality worldwide. Patients with BD have a life expectancy 9 to 17 years lower than that of normal people. BD is a predominant mental disorder, but it can be misdiagnosed as depressive disorder, which leads to difficulties in treating affected patients. Approximately 60% of patients with BD are treated for depression. However, machine learning provides advanced skills and techniques for better diagnosis of BD.

Objective

This review aims to explore the machine learning algorithms used for the detection and diagnosis of bipolar disorder and its subtypes.

Methods

The study protocol adopted the PRISMA-ScR (Preferred Reporting Items for Systematic Reviews and Meta-Analyses Extension for Scoping Reviews) guidelines. We explored 3 databases, namely Google Scholar, ScienceDirect, and PubMed. To enhance the search, we performed backward screening of all the references of the included studies. Based on the predefined selection criteria, 2 levels of screening were performed: title and abstract review, and full review of the articles that met the inclusion criteria. Data extraction was performed independently by all investigators. To synthesize the extracted data, a narrative synthesis approach was followed.

Results

We retrieved 573 potential articles were from the 3 databases. After preprocessing and screening, only 33 articles that met our inclusion criteria were identified. The most commonly used data belonged to the clinical category (19, 58%). We identified different machine learning models used in the selected studies, including classification models (18, 55%), regression models (5, 16%), model-based clustering methods (2, 6%), natural language processing (1, 3%), clustering algorithms (1, 3%), and deep learning–based models (3, 9%). Magnetic resonance imaging data were most commonly used for classifying bipolar patients compared to other groups (11, 34%), whereas microarray expression data sets and genomic data were the least commonly used. The maximum ratio of accuracy was 98%, whereas the minimum accuracy range was 64%.

Conclusions

This scoping review provides an overview of recent studies based on machine learning models used to diagnose patients with BD regardless of their demographics or if they were compared to patients with psychiatric diagnoses. Further research can be conducted to provide clinical decision support in the health industry.

Subtyping Schizophrenia Patients Based on Patterns of Structural Brain Alterations

Schizophrenia is a complex and heterogeneous syndrome. Whether quantitative imaging biomarkers can identify discrete subgroups of patients as might be used to foster personalized medicine approaches for patient care remains unclear. Cross-sectional structural MR images of 163 never-treated first-episode schizophrenia patients (FES) and 133 chronically ill patients with midcourse schizophrenia from the Bipolar and Schizophrenia Network for Intermediate Phenotypes (B-SNIP) consortium and a total of 403 healthy controls were recruited. Morphometric measures (cortical thickness, surface area, and subcortical structures) were extracted for each subject and then the optimized subtyping results were obtained with nonsupervised cluster analysis. Three subgroups of patients defined by distinct patterns of regional cortical and subcortical morphometric features were identified in FES. A similar three subgroup pattern was identified in the independent dataset of patients from the multi-site B-SNIP consortium. Similarities of classification patterns across these two patient cohorts suggest that the 3-group typology is relatively stable over the course of illness. Cognitive functions were worse in subgroup 1 with midcourse schizophrenia than those in subgroup 3. These findings provide novel insight into distinct subgroups of patients with schizophrenia based on structural brain features. Findings of different cognitive functions among the subgroups support clinical differences in the MRI-defined illness subtypes. Regardless of clinical presentation and stage of illness, anatomic MR subgrouping biomarkers can separate neurobiologically distinct subgroups of schizophrenia patients, which represent an important and meaningful step forward in differentiating subtypes of patients for studies of illness neurobiology and potentially for clinical trials.

Artificial intelligence and machine learning-aided drug discovery in central nervous system diseases: State-of-the-arts and future directions

Neurological disorders significantly outnumber diseases in other therapeutic areas. However, developing drugs for central nervous system (CNS) disorders remains the most challenging area in drug discovery, accompanied with the long timelines and high attrition rates. With the rapid growth of biomedical data enabled by advanced experimental technologies, artificial intelligence (AI) and machine learning (ML) have emerged as an indispensable tool to draw meaningful insights and improve decision making in drug discovery. Thanks to the advancements in AI and ML algorithms, now the AI/ML-driven solutions have an unprecedented potential to accelerate the process of CNS drug discovery with better success rate. In this review, we comprehensively summarize AI/ML-powered pharmaceutical discovery efforts and their implementations in the CNS area. After introducing the AI/ML models as well as the conceptualization and data preparation, we outline the applications of AI/ML technologies to several key procedures in drug discovery, including target identification, compound screening, hit/lead generation and optimization, drug response and synergy prediction, de novo drug design, and drug repurposing. We review the current state-of-the-art of AI/ML-guided CNS drug discovery, focusing on blood-brain barrier permeability prediction and implementation into therapeutic discovery for neurological diseases. Finally, we discuss the major challenges and limitations of current approaches and possible future directions that may provide resolutions to these difficulties.

Development and Feasibility Study of an Addiction-Focused Phenotyping Assessment Battery

BACKGROUND AND OBJECTIVES

Current methods of classifying individuals with substance use disorder (SUD) result in vast heterogeneity among persons within a given diagnosis. These approaches, while clinically allowing for distinctions between patient groups, are less than ideal when attempting to recruit a neurobehaviorally defined subset of subjects into clinical trials. To address this gap, alternative strategies have been proposed, including behavioral phenotyping. The NIDA Phenotyping Assessment Battery (PhAB) is a modular package of assessments and neurocognitive tasks that was developed for use in clinical trials. The goal of the present study is to assess the feasibility of the NIDA PhAB with regard to ease of administration and time burden.

METHODS

Healthy controls, persons with cocaine use disorder (CocUD), opioid use disorder (OUD), cannabis use disorder (CanUD), and combined opioid and cocaine use disorder (OCUD) were recruited from various sources (N = 595). Participants completed screening and one to three assessment visits. Time to complete the measures was recorded and a satisfaction interview was administered.

RESULTS

Of the participants enrolled, 381 were deemed eligible. The majority of eligible participants (83%) completed all assessments. The average completion time was 3 hours. High participant satisfaction ratings were noted, with over 90% of participants endorsing a willingness to participate in a similar study and recommend the study to others. CONCLUSION AND SCIENTIFIC SIGNIFICANCE: These findings corroborate the ease with which the PhAB may be easily incorporated into a study assessment visit without undue participant burden. The PhAB is an efficient method for behavioral phenotyping in addiction clinical trials. (Am J Addict 2021;00:00-00).

The Role of Machine Learning in Diagnosing Bipolar Disorder: Scoping Review (Preprint).. [DOI: 10.2196/preprints.29749] [Citation(s) in RCA: 0] [Impact Index Per Article: 0]

BACKGROUND

Bipolar disorder (BD) is the 10th most common cause of frailty in young individuals and has triggered morbidity and mortality worldwide. Patients with BD have a life expectancy 9 to 17 years lower than that of normal people. BD is a predominant mental disorder, but it can be misdiagnosed as depressive disorder, which leads to difficulties in treating affected patients. Approximately 60% of patients with BD are treated for depression. However, machine learning provides advanced skills and techniques for better diagnosis of BD.

OBJECTIVE

This review aims to explore the machine learning algorithms used for the detection and diagnosis of bipolar disorder and its subtypes.

METHODS

The study protocol adopted the PRISMA-ScR (Preferred Reporting Items for Systematic Reviews and Meta-Analyses Extension for Scoping Reviews) guidelines. We explored 3 databases, namely Google Scholar, ScienceDirect, and PubMed. To enhance the search, we performed backward screening of all the references of the included studies. Based on the predefined selection criteria, 2 levels of screening were performed: title and abstract review, and full review of the articles that met the inclusion criteria. Data extraction was performed independently by all investigators. To synthesize the extracted data, a narrative synthesis approach was followed.

RESULTS

We retrieved 573 potential articles were from the 3 databases. After preprocessing and screening, only 33 articles that met our inclusion criteria were identified. The most commonly used data belonged to the clinical category (19, 58%). We identified different machine learning models used in the selected studies, including classification models (18, 55%), regression models (5, 16%), model-based clustering methods (2, 6%), natural language processing (1, 3%), clustering algorithms (1, 3%), and deep learning–based models (3, 9%). Magnetic resonance imaging data were most commonly used for classifying bipolar patients compared to other groups (11, 34%), whereas microarray expression data sets and genomic data were the least commonly used. The maximum ratio of accuracy was 98%, whereas the minimum accuracy range was 64%.

CONCLUSIONS

This scoping review provides an overview of recent studies based on machine learning models used to diagnose patients with BD regardless of their demographics or if they were compared to patients with psychiatric diagnoses. Further research can be conducted to provide clinical decision support in the health industry.

[Psychotic disorders in ICD-11: the revisions]

Psychotic disorders in ICD-11: the revisions Abstract. This article provides an overview of the main changes to the chapter "Schizophrenia or Other Primary Psychotic Disorders" (6A2) from ICD-10 to ICD-11 and compares them with the psychosis chapter of DSM-5. These changes include abandoning the classical subtypes of Schizophrenia as well as of the special significance of Schneider's first-rank symptoms, resulting in the general requirement of two key features (one must be a positive symptom) in the definition of "Schizophrenia" (6A20) and the allowance for bizarre contents in "Delusional Disorder" (6A24), which now includes "Induced Delusional Disorder" (F24). Further introduced are the focus on the current episode, the restriction of "Acute and Transient Psychotic Disorder" (6A23) to the former Polymorphic Disorder Without Schizophrenic Symptoms (F23.0), the diagnosis of delusional "Obsessive-Compulsive or Related Disorders" (6B2) exclusively as Obsessive-Compulsive Disorders, the specification of "Schizoaffective Disorder" (6A21), and the formulation of a distinct subchapter "Catatonia" (6A4) for the assessment of catatonic features in the context of several disorders. In analogy to DSM-5, ICD-11 now includes the optional category "Symptomatic Manifestations of Primary Psychotic Disorders" (6A25) for the dimensional quantification of symptoms. Again, developmental aspects remain unattended in in the ICD-11-definitions of psychotic disorders.

[Brain imaging in schizophrenia : A review of current trends and developments]

Imaging methods have become the main approach for identifying dysfunctional neuronal networks in schizophrenia. This review article presents recent results of disorders of neuronal networks at structural and functional levels and summarizes the current developments. Large multicenter analyses have further established patterns of regional brain alterations, while novel methods in magnetic resonance (MR) morphometry have contributed to differentiating early from delayed brain structural changes. The use of machine learning approaches has not only enabled the establishment of classification models using biological data for future differential diagnostic use, it has also facilitated multivariate models for outcome prediction following therapeutic interventions. Novel methods, such as BrainAGE, a surrogate marker of accelerated brain aging processes, have added to longitudinal studies to gain insights into the brain structural dynamics from early brain developmental alterations to progressive structural brain changes after disease onset.

Using machine learning to explain the heterogeneity of schizophrenia. Realizing the promise and avoiding the hype

Despite extensive research and prodigious advances in neuroscience, our comprehension of the nature of schizophrenia remains rudimentary. Our failure to make progress is attributed to the extreme heterogeneity of this condition, enormous complexity of the human brain, limitations of extant research paradigms, and inadequacy of traditional statistical methods to integrate or interpret increasingly large amounts of multidimensional information relevant to unravelling brain function. Fortunately, the rapidly developing science of machine learning appears to provide tools capable of addressing each of these impediments. Enthusiasm about the potential of machine learning methods to break the current impasse is reflected in the steep increase in the number of scientific publication about the application of machine learning to the study of schizophrenia. Machine learning approaches are, however, poorly understood by schizophrenia researchers and clinicians alike. In this paper, we provide a simple description of the nature and techniques of machine learning and their application to the study of schizophrenia. We then summarize its potential and constraints with illustrations from six studies of machine learning in schizophrenia and address some common misconceptions about machine learning. We suggest some guidelines for researchers, readers, science editors and reviewers of the burgeoning machine learning literature in schizophrenia. In order to realize its enormous promise, we suggest the need for the disciplined application of machine learning methods to the study of schizophrenia with a clear recognition of its capability and challenges accompanied by a concurrent effort to improve machine learning literacy among neuroscientists and mental health professionals.

Imaging-Based Subtyping for Psychiatric Syndromes

Despite considerable research evidence demonstrating significant neurobiological alterations in psychiatric disorders, incorporating neuroimaging approaches into clinical practice remains challenging. There is an urgent need for biologically validated psychiatric disease constructs that can inform diagnostic algorithms and targeted treatment development. In this article, we present a conceptual review of the most robust and impactful findings from studies that use neuroimaging methods in efforts to define distinct disease subtypes, while emphasizing cross-diagnostic and dimensional approaches. In addition, we discuss current challenges in psychoradiology and outline potential future strategies for clinically applicable translation.

Identifying psychosis spectrum disorder from experience sampling data using machine learning approaches

The ubiquity of smartphones opened up the possibility of widespread use of the Experience Sampling Method (ESM). The method is used to collect longitudinal data of participants' daily life experiences and is ideal to capture fluctuations in emotions (momentary mental states) as an indicator for later mental ill-health. In this study, ESM data of patients with psychosis spectrum disorder and controls were used to examine daily life emotions and higher order patterns thereof. We attempted to determine whether aggregated ESM data, in which statistical measures represent the distribution and dynamics of the original data, were able to distinguish patients from controls in a predictive modeling framework. Variable importance, recursive feature elimination, and ReliefF methods were used for feature selection. Model training, tuning, and testing were performed in nested cross-validation, based on algorithms such as Random Forests, Support Vector Machines, Gaussian Processes, Logistic Regression, and Neural Networks. ROC analysis was used to post-process these models. Stability of model performance was studied using Monte Carlo simulations. The results provide evidence that patterns in emotion changes can be captured by applying a combination of these techniques. Acceleration in the variables anxious and insecure was particularly successful in adding further predictive power to the models. The best results were achieved by Support Vector Machines with radial kernel (accuracy = 82% and sensitivity = 82%). This proof-of-concept work demonstrates that synergistic machine learning and statistical modeling may be used to harness the power of ESM data in the future.