Identifying schizophrenia subgroups using clustering and supervised learning

Connectional-style-guided contextual representation learning for brain disease diagnosis

Structural magnetic resonance imaging (sMRI) has shown great clinical value and has been widely used in deep learning (DL) based computer-aided brain disease diagnosis. Previous DL-based approaches focused on local shapes and textures in brain sMRI that may be significant only within a particular domain. The learned representations are likely to contain spurious information and have poor generalization ability in other diseases and datasets. To facilitate capturing meaningful and robust features, it is necessary to first comprehensively understand the intrinsic pattern of the brain that is not restricted within a single data/task domain. Considering that the brain is a complex connectome of interlinked neurons, the connectional properties in the brain have strong biological significance, which is shared across multiple domains and covers most pathological information. In this work, we propose a connectional style contextual representation learning model (CS-CRL) to capture the intrinsic pattern of the brain, used for multiple brain disease diagnosis. Specifically, it has a vision transformer (ViT) encoder and leverages mask reconstruction as the proxy task and Gram matrices to guide the representation of connectional information. It facilitates the capture of global context and the aggregation of features with biological plausibility. The results indicate that CS-CRL achieves superior accuracy in multiple brain disease diagnosis tasks across six datasets and three diseases and outperforms state-of-the-art models. Furthermore, we demonstrate that CS-CRL captures more brain-network-like properties, and better aggregates features, is easier to optimize, and is more robust to noise, which explains its superiority in theory.

Mapping Brain Synergy Dysfunction in Schizophrenia: Understanding Individual Differences and Underlying Molecular Mechanisms

To elucidate the brain-wide information interactions that vary and contribute to individual differences in schizophrenia (SCZ), an information-resolved method is employed to construct individual synergistic and redundant interaction matrices based on regional pairwise BOLD time-series from 538 SCZ and 540 normal controls (NC). This analysis reveals a stable pattern of regionally-specific synergy dysfunction in SCZ. Furthermore, a hierarchical Bayesian model is applied to deconstruct the patterns of whole-brain synergy dysfunction into three latent factors that explain symptom heterogeneity in SCZ. Factor 1 exhibits a significant positive correlation with Positive and Negative Syndrome Scale (PANSS) positive scores, while factor 3 demonstrates significant negative correlations with PANSS negative and general scores. By integrating the neuroimaging data with normative gene expression information, this study identifies that each of these three factors corresponded to a subset of the SCZ risk gene set. Finally, by combining data from NeuroSynth and open molecular imaging sources, along with a spatially heterogeneous mean-field model, this study delineates three SCZ synergy factors corresponding to distinct symptom profiles and implicating unique cognitive, neurodynamic, and neurobiological mechanisms.

Data-driven multivariate identification of gyrification patterns in a transdiagnostic patient cohort: A cluster analysis approach

BACKGROUND

Multivariate data-driven statistical approaches offer the opportunity to study multi-dimensional interdependences between a large set of biological parameters, such as high-dimensional brain imaging data. For gyrification, a putative marker of early neurodevelopment, direct comparisons of patterns among multiple psychiatric disorders and investigations of potential heterogeneity of gyrification within one disorder and a transdiagnostic characterization of neuroanatomical features are lacking.

METHODS

In this study we used a data-driven, multivariate statistical approach to analyze cortical gyrification in a large cohort of N = 1028 patients with major psychiatric disorders (Major depressive disorder: n = 783, bipolar disorder: n = 129, schizoaffective disorder: n = 44, schizophrenia: n = 72) to identify cluster patterns of gyrification beyond diagnostic categories.

RESULTS

Cluster analysis applied on gyrification data of 68 brain regions (DK-40 atlas) identified three clusters showing difference in overall (global) gyrification and minor regional variation (regions). Newly, data-driven subgroups are further discriminative in cognition and transdiagnostic disease risk factors.

CONCLUSIONS

Results indicate that gyrification is associated with transdiagnostic risk factors rather than diagnostic categories and further imply a more global role of gyrification related to mental health than a disorder specific one. Our findings support previous studies highlighting the importance of association cortices involved in psychopathology. Explorative, data-driven approaches like ours can help to elucidate if the brain imaging data on hand and its a priori applied grouping actually has the potential to find meaningful effects or if previous hypotheses about the phenotype as well as its grouping have to be revisited.

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.

Characterization of early psychosis patients carrying a genetic vulnerability to redox dysregulation: a computational analysis of mechanism-based gene expression profile in fibroblasts

In view of its heterogeneity, schizophrenia needs new diagnostic tools based on mechanistic biomarkers that would allow early detection. Complex interaction between genetic and environmental risk factors may lead to NMDAR hypofunction, inflammation and redox dysregulation, all converging on oxidative stress. Using computational analysis, the expression of 76 genes linked to these systems, known to be abnormally regulated in schizophrenia, was studied in skin-fibroblasts from early psychosis patients and age-matched controls (N = 30), under additional pro-oxidant challenge to mimic environmental stress. To evaluate the contribution of a genetic risk related to redox dysregulation, we investigated the GAG trinucleotide polymorphism in the key glutathione (GSH) synthesizing enzyme, glutamate-cysteine-ligase-catalytic-subunit (gclc) gene, known to be associated with the disease. Patients and controls showed different gene expression profiles that were modulated by GAG-gclc genotypes in combination with oxidative challenge. In GAG-gclc low-risk genotype patients, a global gene expression dysregulation was observed, especially in the antioxidant system, potentially induced by other risks. Both controls and patients with GAG-gclc high-risk genotype (gclcGAG-HR) showed similar gene expression profiles. However, under oxidative challenge, a boosting of other antioxidant defense, including the master regulator Nrf2 and TRX systems was observed only in gclcGAG-HR controls, suggesting a protective compensation against the genetic GSH dysregulation. Moreover, RAGE (redox/inflammation interaction) and AGMAT (arginine pathway) were increased in the gclcGAG-HR patients, suggesting some additional risk factors interacting with this genotype. Finally, the use of a machine-learning approach allowed discriminating patients and controls with an accuracy up to 100%, paving the way towards early detection of schizophrenia.

Cortical thinning in relation to impaired insight into illness in patients with treatment resistant schizophrenia

Impaired insight into illness is a common element of schizophrenia that contributes to treatment nonadherence and negative clinical outcomes. Previous studies suggest that impaired insight may arise from brain abnormalities. However, interpretations of these findings are limited due to small sample sizes and inclusion of patients with a narrow range of illness severity and insight deficits. In a large sample of patients with schizophrenia, the majority of which were designated as treatment-resistant, we investigated the associations between impaired insight and cortical thickness and subcortical volumes. A total of 94 adult participants with a schizophrenia spectrum disorder were included. Fifty-six patients (60%) had treatment-resistant schizophrenia. The core domains of insight were assessed with the VAGUS insight into psychosis scale. We obtained 3T MRI T1-weighted images, which were analysed using CIVET and MAGeT-Brain. Whole-brain vertex-wise analyses revealed impaired insight, as measured by VAGUS average scores, was related to cortical thinning in left frontotemporoparietal regions. The same analysis in treatment-resistant patients showed thinning in the same regions, even after controlling for age, sex, illness severity, and chlorpromazine antipsychotic dose equivalents. No association was found in non-treatment-resistant patients. Region-of-interest analyses revealed impaired general illness awareness was associated with cortical thinning in the left supramarginal gyrus when controlling for covariates. Reduced right and left thalamic volumes were associated with VAGUS symptom attribution and awareness of negative consequences subscale scores, respectively, but not after correction for multiple testing. Our results suggest impaired insight into illness is related to cortical thinning in left frontotemporoparietal regions in patients with schizophrenia, particularly those with treatment resistance where insight deficits may be more chronic.

One Size Does Not Fit All: Methodological Considerations for Brain-Based Predictive Modeling in Psychiatry

Psychiatric illnesses are heterogeneous in nature. No illness manifests in the same way across individuals, and no two patients with a shared diagnosis exhibit identical symptom profiles. Over the last several decades, group-level analyses of in vivo neuroimaging data have led to fundamental advances in our understanding of the neurobiology of psychiatric illnesses. More recently, access to computational resources and large, publicly available datasets alongside the rise of predictive modeling and precision medicine approaches have facilitated the study of psychiatric illnesses at an individual level. Data-driven machine learning analyses can be applied to identify disease-relevant biological subtypes, predict individual symptom profiles, and recommend personalized therapeutic interventions. However, when developing these predictive models, methodological choices must be carefully considered to ensure accurate, robust, and interpretable results. Choices pertaining to algorithms, neuroimaging modalities and states, data transformation, phenotypes, parcellations, sample sizes, and populations we are specifically studying can influence model performance. Here, we review applications of neuroimaging-based machine learning models to study psychiatric illnesses and discuss the effects of different methodological choices on model performance. An understanding of these effects is crucial for the proper implementation of predictive models in psychiatry and will facilitate more accurate diagnoses, prognoses, and therapeutics.

Intra and inter-individual variability in functional connectomes of patients with First Episode of Psychosis

Patients with Schizophrenia may show different clinical presentations, not only regarding inter-individual comparisons but also in one specific subject over time. In fMRI studies, functional connectomes have been shown to carry valuable individual level information, which can be associated with cognitive and behavioral variables. Moreover, functional connectomes have been used to identify subjects within a group, as if they were fingerprints. For the particular case of Schizophrenia, it has been shown that there is reduced connectome stability as well as higher inter-individual variability. Here, we studied inter and intra-individual heterogeneity by exploring functional connectomes' variability and related it with clinical variables (PANSS Total scores and antipsychotic's doses). Our sample consisted of 30 patients with First Episode of Psychosis and 32 Healthy Controls, with a test-retest approach of two resting-state fMRI scanning sessions. In our patients' group, we found increased deviation from healthy functional connectomes and increased intragroup inter-subject variability, which was positively correlated to symptoms' levels in six subnetworks (visual, somatomotor, dorsal attention, ventral attention, frontoparietal and DMN). Moreover, changes in symptom severity were positively related to changes in deviation from healthy functional connectomes. Regarding intra-subject variability, we were unable to replicate previous findings of reduced connectome stability (i.e., increased intra-subject variability), but we found a trend suggesting that result. Our findings highlight the relevance of variability characterization in Schizophrenia, and they can be related to evidence of Schizophrenia patients having a noisy functional connectome.

Cluster-based psychological phenotyping and differences in anxiety treatment outcomes

The identification of markers of mental health illness treatment response and susceptibility using personalized medicine has been elusive. In the context of psychological treatment for anxiety, we conducted two studies to identify psychological phenotypes with distinct characteristics related to: psychological intervention modalities (mindfulness training/awareness), mechanism of action (worry), and clinical outcome (generalized anxiety disorder scale scores). We also examined whether phenotype membership interacted with treatment response (Study 1) and mental health illness diagnosis (Studies 1-2). Interoceptive awareness, emotional reactivity, worry, and anxiety were assessed at baseline in treatment-seeking individuals (Study 1, n = 63) and from the general population (Study 2, n = 14,010). In Study 1, participants were randomly assigned to an app-delivered mindfulness program for anxiety for two months or treatment as usual. Changes in anxiety were assessed 1 and 2 months post-treatment initiation. In studies 1-2, three phenotypes were identified: 'severely anxious with body/emotional awareness' (cluster 1), 'body/emotionally unaware' (cluster 2), and 'non-reactive and aware' (cluster 3). Study 1's results revealed a significant treatment response relative to controls (ps < 0.001) for clusters 1 and 3, but not for cluster 2. Chi-square analyses revealed that phenotypes exhibited significantly different proportions of participants with mental health diagnoses (studies 1-2). These results suggest that psychological phenotyping can bring the application of personalized medicine into clinical settings.Registry name and URL: Developing a novel digital therapeutic for the treatment of generalized anxiety disorder https://clinicaltrials.gov/ct2/show/NCT03683472?term=judson+brewer&draw=1&rank=1 .Trial registration: Registered at clinicaltrials.gov (NCT03683472) on 25/09/2018.

Application of machine learning methods in predicting schizophrenia and bipolar disorders: A systematic review

Background and Aim

Schizophrenia and bipolar disorder (BD) are critical and high-risk inherited mental disorders with debilitating symptoms. Worldwide, 3% of the population suffers from these disorders. The mortality rate of these patients is higher compared to other people. Current procedures cannot effectively diagnose these disorders because it takes an average of 10 years from the onset of the first symptoms to the definitive diagnosis of the disease. Machine learning (ML) techniques are used to meet this need. This study aimed to summarize information on the use of ML techniques for predicting schizophrenia and BD to help early and timely diagnosis of the disease.

Methods

A systematic literature search included articles published until January 19, 2020 in 3 databases. Two reviewers independently assessed original papers to determine eligibility for inclusion in this review. PRISMA guidelines were followed to conduct the study, and the Prediction Model Risk of Bias Assessment Tool (PROBAST) to assess included papers.

Results

In this review, 1243 papers were retrieved through database searches, of which 15 papers were included based on full-text assessment. ML techniques were used to predict schizophrenia and BDs. The main algorithms applied were support vector machine (SVM) (10 studies), random forests (RF) (5 studies), and gradient boosting (GB) (3 studies). Input and output characteristics were very diverse and have been kept to enable future research. RFs algorithms demonstrated significantly higher accuracy and sensitivity than SVM and GB. GB demonstrated significantly higher specificity than SVM and RF. We found no significant difference between RF and SVM in terms of specificity.

Conclusion

ML can precisely predict results and assist in making clinical decisions-concerning schizophrenia and BD. RF often performed better than other algorithms in supervised learning tasks. This study identified gaps in the literature and opportunities for future psychological ML research.

Deconstructing heterogeneity in schizophrenia through language: a semi-automated linguistic analysis and data-driven clustering approach

Previous works highlighted the relevance of automated language analysis for predicting diagnosis in schizophrenia, but a deeper language-based data-driven investigation of the clinical heterogeneity through the illness course has been generally neglected. Here we used a semiautomated multidimensional linguistic analysis innovatively combined with a machine-driven clustering technique to characterize the speech of 67 individuals with schizophrenia. Clusters were then compared for psychopathological, cognitive, and functional characteristics. We identified two subgroups with distinctive linguistic profiles: one with higher fluency, lower lexical variety but greater use of psychological lexicon; the other with reduced fluency, greater lexical variety but reduced psychological lexicon. The former cluster was associated with lower symptoms and better quality of life, pointing to the existence of specific language profiles, which also show clinically meaningful differences. These findings highlight the importance of considering language disturbances in schizophrenia as multifaceted and approaching them in automated and data-driven ways.

Greater individual variability in functional brain activity during working memory performance in Schizophrenia Spectrum Disorders (SSD)

Heterogeneity has been a persistent challenge in understanding Schizophrenia Spectrum Disorders (SSD). Traditional case-control comparisons often show variable results, and may not map well onto individuals. To better understand heterogeneity and group differences in SSD compared to typically developing controls (TDC), we examined variability in functional brain activity during a working memory (WM) task with known deficits in SSD. Neuroimaging and behavioural data were extracted from two datasets collectively providing 34 TDC and 56 individuals with SSD (n = 90). Functional activity in response to an N-Back WM task (3-Back vs 1-Back) was examined between and within groups. Individual variability was calculated via the correlational distance of fMRI activity maps between participants; mean correlational distance from one participant to all others was defined as a 'variability score'. Greater individual variability in functional activity was found in SSD compared to TDC (p = 0.00090). At the group level, a case-control comparison suggested SSD had reduced activity in task positive and task negative networks. However, when SSD were divided into high and low variability subgroups, the low variability groups showed no differences relative to TDC while the high variability group showed little activity at the group level. Our results imply prior case-control differences may be driven by a subgroup of SSD who do not show specific impairments but instead show more 'idiosyncratic' activity patterns. In SSD but not TDC, variability was also related to cognitive performance and age. This novel approach focusing on individual variability has important implications for understanding the neurobiology of SSD.

Two subtypes of schizophrenia identified by an individual-level atypical pattern of tensor-based morphometric measurement

Difficulties in parsing the multiaspect heterogeneity of schizophrenia (SCZ) based on current nosology highlight the need to subtype SCZ using objective biomarkers. Here, utilizing a large-scale multisite SCZ dataset, we identified and validated 2 neuroanatomical subtypes with individual-level abnormal patterns of the tensor-based morphometric measurement. Remarkably, compared with subtype 1, which showed moderate deficits of some subcortical nuclei and an enlarged striatum and cerebellum, subtype 2, which showed cerebellar atrophy and more severe subcortical nuclei atrophy, had a higher subscale score of negative symptoms, which is considered to be a core aspect of SCZ and is associated with functional outcome. Moreover, with the neuroimaging-clinic association analysis, we explored the detailed relationship between the heterogeneity of clinical symptoms and the heterogeneous abnormal neuroanatomical patterns with respect to the 2 subtypes. And the neuroimaging-transcription association analysis highlighted several potential heterogeneous biological factors that may underlie the subtypes. Our work provided an effective framework for investigating the heterogeneity of SCZ from multilevel aspects and may provide new insights for precision psychiatry.

Deep learning-based integration of genetics with registry data for stratification of schizophrenia and depression

Currently, psychiatric diagnoses are, in contrast to most other medical fields, based on subjective symptoms and observable signs and call for new and improved diagnostics to provide the most optimal care. On the basis of a deep learning approach, we performed unsupervised patient stratification of 19,636 patients with depression [major depressive disorder (MDD)] and/or schizophrenia (SCZ) and 22,467 population controls from the iPSYCH2012 case cohort. We integrated data of disorder severity, history of mental disorders and disease comorbidities, genetics, and medical birth data. From this, we stratified the individuals in six and seven unique clusters for MDD and SCZ, respectively. When censoring data until diagnosis, we could predict MDD clusters with areas under the curve (AUCs) of 0.54 to 0.80 and SCZ clusters with AUCs of 0.71 to 0.86. Overall cases and controls could be predicted with an AUC of 0.81, illustrating the utility of data-driven subgrouping in psychiatry.

Discriminative Analysis of Schizophrenia Patients Using Topological Properties of Structural and Functional Brain Networks: A Multimodal Magnetic Resonance Imaging Study

Interest in the application of machine learning (ML) techniques to multimodal magnetic resonance imaging (MRI) data for the diagnosis of schizophrenia (SZ) at the individual level is growing. However, a few studies have applied the features of structural and functional brain networks derived from multimodal MRI data to the discriminative analysis of SZ patients at different clinical stages. In this study, 205 normal controls (NCs), 61 first-episode drug-naive SZ (FESZ) patients, and 79 chronic SZ (CSZ) patients were recruited. We acquired their structural MRI, diffusion tensor imaging, and resting-state functional MRI data and constructed brain networks for each participant, including the gray matter network (GMN), white matter network (WMN), and functional brain network (FBN). We then calculated 3 nodal properties for each brain network, including degree centrality, nodal efficiency, and betweenness centrality. Two classifications (SZ vs. NC and FESZ vs. CSZ) were performed using five ML algorithms. We found that the SVM classifier with the input features of the combination of nodal properties of both the GMN and FBN achieved the best performance to discriminate SZ patients from NCs [accuracy, 81.2%; area under the receiver operating characteristic curve (AUC), 85.2%; p &lt; 0.05]. Moreover, the SVM classifier with the input features of the combination of the nodal properties of both the GMN and WMN achieved the best performance to discriminate FESZ from CSZ patients (accuracy, 86.2%; AUC, 92.3%; p &lt; 0.05). Furthermore, the brain areas in the subcortical/cerebellum network and the frontoparietal network showed significant importance in both classifications. Together, our findings provide new insights to understand the neuropathology of SZ and further highlight the potential advantages of multimodal network properties for identifying SZ patients at different clinical stages.

A Cross-Sectional Study on Associations Between BDNF, CRP, IL-6 and Clinical Symptoms, Cognitive and Personal Performance in Patients With Paranoid Schizophrenia

BACKGROUND

Cognitive impairment is among the core dimensions in schizophrenia and is a significant predictor of everyday functioning in people with schizophrenia. Given the enormous burden of schizophrenia, the search for its clinically relevant biomarkers is essential. Researchers have been trying to elucidate factors of cognitive impairment as well as personal performance, but the search is still ongoing. The aim of the study was to search for associations between BDNF, CRP, IL-6 and clinical symptoms, cognitive and personal performance in patients with paranoid schizophrenia.

METHODS

A total of 86 patients (53.5% women, mean age 31.1 ± 6.5) with paranoid schizophrenia (F20.0; ICD-10) in remission were examined. Clinical and neuropsychological examination included the Positive and Negative Syndrome Scale, Personal and Social Performance Scale, Calgary Depression Scale for Schizophrenia and the Brief Assessment of Cognitive Function in Schizophrenia. IL-6, BDNF, CRP levels were determined in the patients' blood serum.

RESULTS

Cognitive impairment was revealed in 79.1% of patients and was more profound in patients with higher number of hospitalizations (p = 0.006). The average BDNF levels were 13.38 ± 15.84 ng/ml, CRP concentration was 2.09 ± 2.54 mg/l, and IL-6 levels were 12.14 ± 5.88 pg/ml. There were no differences in biomarker levels or BACS results in patients that had different antipsychotic therapy or differed in the presence of anticholinergic therapy. CRP levels were higher in patients with longer disease duration, lower age of onset, more impaired personal social performance and processing speed. IL-6 was higher in individuals with lower working memory scores. PANSS negative subscale score negatively correlated and PSP score positively correlated with most cognitive domains. A linear regression established that the first episode vs. multiple episodes of schizophrenia could statistically significantly predict personal and social performance and cognition, including speech fluency and planning, as well as CRP levels.

CONCLUSIONS

This study continues the search for biomarkers of schizophrenia and cognitive impairment in schizophrenia to improve the reliability of diagnosing the disorder and find new treatment approaches. The role of the number of psychoses experienced (first episode vs. multiple episodes of schizophrenia) in cognition, personal and social performance and inflammation is shown.

Prediction of Early Symptom Remission in Two Independent Samples of First-Episode Psychosis Patients Using Machine Learning

BACKGROUND

Validated clinical prediction models of short-term remission in psychosis are lacking. Our aim was to develop a clinical prediction model aimed at predicting 4-6-week remission following a first episode of psychosis.

METHOD

Baseline clinical data from the Athens First Episode Research Study was used to develop a Support Vector Machine prediction model of 4-week symptom remission in first-episode psychosis patients using repeated nested cross-validation. This model was further tested to predict 6-week remission in a sample of two independent, consecutive Danish first-episode cohorts.

RESULTS

Of the 179 participants in Athens, 120 were male with an average age of 25.8 years and average duration of untreated psychosis of 32.8 weeks. 62.9% were antipsychotic-naïve. Fifty-seven percent attained remission after 4 weeks. In the Danish cohort, 31% attained remission. Eleven clinical scale items were selected in the Athens 4-week remission cohort. These included the Duration of Untreated Psychosis, Personal and Social Performance Scale, Global Assessment of Functioning and eight items from the Positive and Negative Syndrome Scale. This model significantly predicted 4-week remission status (area under the receiver operator characteristic curve (ROC-AUC) = 71.45, P < .0001). It also predicted 6-week remission status in the Danish cohort (ROC-AUC = 67.74, P < .0001), demonstrating reliability.

CONCLUSIONS

Using items from common and validated clinical scales, our model significantly predicted early remission in patients with first-episode psychosis. Although replicated in an independent cohort, forward testing between machine learning models and clinicians' assessment should be undertaken to evaluate the possible utility as a routine clinical tool.

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.

Identification of minimal hepatic encephalopathy based on dynamic functional connectivity

To investigate whether dynamic functional connectivity (DFC) metrics can better identify minimal hepatic encephalopathy (MHE) patients from cirrhotic patients without any hepatic encephalopathy (noHE) and healthy controls (HCs). Resting-state functional MRI data were acquired from 62 patients with cirrhosis (MHE, n = 30; noHE, n = 32) and 41 HCs. We used the sliding time window approach and functional connectivity analysis to extract the time-varying properties of brain connectivity. Three DFC characteristics (i.e., strength, stability, and variability) were calculated. For comparison, we also calculated the static functional connectivity (SFC). A linear support vector machine was used to differentiate MHE patients from noHE and HCs using DFC and SFC metrics as classification features. The leave-one-out cross-validation method was used to estimate the classification performance. The strength of DFC (DFC-Dstrength) achieved the best accuracy (MHE vs. noHE, 72.5%; MHE vs. HCs, 84%; and noHE vs. HCs, 88%) compared to the other dynamic features. Compared to static features, the classification accuracies of the DFC-Dstrength feature were improved by 10.5%, 8%, and 14% for MHE vs. noHE, MHE vs. HC, and noHE vs. HCs, respectively. Based on the DFC-Dstrength, seven nodes were identified as the most discriminant features to classify MHE from noHE, including left inferior parietal lobule, left supramarginal gyrus, left calcarine, left superior frontal gyrus, left cerebellum, right postcentral gyrus, and right insula. In summary, DFC characteristics have a higher classification accuracy in identifying MHE from cirrhosis patients. Our findings suggest the usefulness of DFC in capturing neural processes and identifying disease-related biomarkers important for MHE identification.

Promises and pitfalls of deep neural networks in neuroimaging-based psychiatric research

By promising more accurate diagnostics and individual treatment recommendations, deep neural networks and in particular convolutional neural networks have advanced to a powerful tool in medical imaging. Here, we first give an introduction into methodological key concepts and resulting methodological promises including representation and transfer learning, as well as modelling domain-specific priors. After reviewing recent applications within neuroimaging-based psychiatric research, such as the diagnosis of psychiatric diseases, delineation of disease subtypes, normative modeling, and the development of neuroimaging biomarkers, we discuss current challenges. This includes for example the difficulty of training models on small, heterogeneous and biased data sets, the lack of validity of clinical labels, algorithmic bias, and the influence of confounding variables.

Artificial intelligence for brain diseases: A systematic review

Artificial intelligence (AI) is a major branch of computer science that is fruitfully used for analyzing complex medical data and extracting meaningful relationships in datasets, for several clinical aims. Specifically, in the brain care domain, several innovative approaches have achieved remarkable results and open new perspectives in terms of diagnosis, planning, and outcome prediction. In this work, we present an overview of different artificial intelligent techniques used in the brain care domain, along with a review of important clinical applications. A systematic and careful literature search in major databases such as Pubmed, Scopus, and Web of Science was carried out using "artificial intelligence" and "brain" as main keywords. Further references were integrated by cross-referencing from key articles. 155 studies out of 2696 were identified, which actually made use of AI algorithms for different purposes (diagnosis, surgical treatment, intra-operative assistance, and postoperative assessment). Artificial neural networks have risen to prominent positions among the most widely used analytical tools. Classic machine learning approaches such as support vector machine and random forest are still widely used. Task-specific algorithms are designed for solving specific problems. Brain images are one of the most used data types. AI has the possibility to improve clinicians' decision-making ability in neuroscience applications. However, major issues still need to be addressed for a better practical use of AI in the brain. To this aim, it is important to both gather comprehensive data and build explainable AI algorithms.

Do neurobiological differences exist between paranoid and non-paranoid schizophrenia? Findings from the bipolar schizophrenia network on intermediate phenotypes study

Subtypes of schizophrenia, constructed using clinical phenomenology to resolve illness heterogeneity, have faced criticism due to overlapping symptomatology and longitudinal instability; they were therefore dropped from the Diagnostic Statistical Manual-5. Cognitive and imaging findings comparing paranoid (P-SZ) and non-paranoid (disorganized, residual and undifferentiated; NP-SZ) schizophrenia have been limited due to small sample sizes. We assessed P-SZ and NP-SZ using symptomatology, cognition and brain structure and predicted that there would be few neurobiological differences. P-SZ (n = 237), NP-SZ (n = 127) and controls (n = 430) were included from a multi-site study. In a subset of this sample, structural imaging measures (P-SZ, n = 133; NP-SZ, n = 67; controls, n = 310) were calculated using Freesurfer 6.0. Group contrasts were run using analysis of covariance, controlling for age, sex, race and site, p-values were corrected using False Discovery Rate (FDR) and were repeated excluding the residual subtype. Compared to NP-SZ (with and without the residual subtype), P-SZ displayed fewer negative symptoms, faster speed of processing, larger bilateral hippocampus, right amygdala and their subfield volumes. Additionally, NP-SZ (with residual subtype) displayed fewer depressive symptoms and higher left transverse temporal cortical thickness (CT) but NP-SZ without residual subtype showed lower GAF scores and worse digit sequencing compared to P-SZ. No differences in positive symptoms and functioning (global or social) were detected. Subtle but significant differences were seen in cognition, symptoms, CT and subcortical volumes between P-SZ and NP-SZ. While the magnitude of these differences is not large enough to justify them as distinct categories, the paranoid- nonparanoid distinction in schizophrenia merits further investigation.

A systematic review and narrative synthesis of data-driven studies in schizophrenia symptoms and cognitive deficits

To tackle the phenotypic heterogeneity of schizophrenia, data-driven methods are often applied to identify subtypes of its symptoms and cognitive deficits. However, a systematic review on this topic is lacking. The objective of this review was to summarize the evidence obtained from longitudinal and cross-sectional data-driven studies in positive and negative symptoms and cognitive deficits in patients with schizophrenia spectrum disorders, their unaffected siblings and healthy controls or individuals from general population. Additionally, we aimed to highlight methodological gaps across studies and point out future directions to optimize the translatability of evidence from data-driven studies. A systematic review was performed through searching PsycINFO, PubMed, PsycTESTS, PsycARTICLES, SCOPUS, EMBASE and Web of Science electronic databases. Both longitudinal and cross-sectional studies published from 2008 to 2019, which reported at least two statistically derived clusters or trajectories were included. Two reviewers independently screened and extracted the data. In this review, 53 studies (19 longitudinal and 34 cross-sectional) that conducted among 17,822 patients, 8729 unaffected siblings and 5520 controls or general population were included. Most longitudinal studies found four trajectories that characterized by stability, progressive deterioration, relapsing and progressive amelioration of symptoms and cognitive function. Cross-sectional studies commonly identified three clusters with low, intermediate (mixed) and high psychotic symptoms and cognitive profiles. Moreover, identified subgroups were predicted by numerous genetic, sociodemographic and clinical factors. Our findings indicate that schizophrenia symptoms and cognitive deficits are heterogeneous, although methodological limitations across studies are observed. Identified clusters and trajectories along with their predictors may be used to base the implementation of personalized treatment and develop a risk prediction model for high-risk individuals with prodromal symptoms.

Challenges and Future Prospects of Precision Medicine in Psychiatry

Precision medicine is increasingly recognized as a promising approach to improve disease treatment, taking into consideration the individual clinical and biological characteristics shared by specific subgroups of patients. In specific fields such as oncology and hematology, precision medicine has already started to be implemented in the clinical setting and molecular testing is routinely used to select treatments with higher efficacy and reduced adverse effects. The application of precision medicine in psychiatry is still in its early phases. However, there are already examples of predictive models based on clinical data or combinations of clinical, neuroimaging and biological data. While the power of single clinical predictors would remain inadequate if analyzed only with traditional statistical approaches, these predictors are now increasingly used to impute machine learning models that can have adequate accuracy even in the presence of relatively small sample size. These models have started to be applied to disentangle relevant clinical questions that could lead to a more effective management of psychiatric disorders, such as prediction of response to the mood stabilizer lithium, resistance to antidepressants in major depressive disorder or stratification of the risk and outcome prediction in schizophrenia. In this narrative review, we summarized the most important findings in precision medicine in psychiatry based on studies that constructed machine learning models using clinical, neuroimaging and/or biological data. Limitations and barriers to the implementation of precision psychiatry in the clinical setting, as well as possible solutions and future perspectives, will be presented.