Modeling essential connections in obsessive-compulsive disorder patients using functional MRI

Dissecting Psychiatric Heterogeneity and Comorbidity with Core Region-Based Machine Learning

Machine learning approaches are increasingly being applied to neuroimaging data from patients with psychiatric disorders to extract brain-based features for diagnosis and prognosis. The goal of this review is to discuss recent practices for evaluating machine learning applications to obsessive-compulsive and related disorders and to advance a novel strategy of building machine learning models based on a set of core brain regions for better performance, interpretability, and generalizability. Specifically, we argue that a core set of co-altered brain regions (namely 'core regions') comprising areas central to the underlying psychopathology enables the efficient construction of a predictive model to identify distinct symptom dimensions/clusters in individual patients. Hypothesis-driven and data-driven approaches are further introduced showing how core regions are identified from the entire brain. We demonstrate a broadly applicable roadmap for leveraging this core set-based strategy to accelerate the pursuit of neuroimaging-based markers for diagnosis and prognosis in a variety of psychiatric disorders.

Sampling inequalities affect generalization of neuroimaging-based diagnostic classifiers in psychiatry

BACKGROUND

The development of machine learning models for aiding in the diagnosis of mental disorder is recognized as a significant breakthrough in the field of psychiatry. However, clinical practice of such models remains a challenge, with poor generalizability being a major limitation.

METHODS

Here, we conducted a pre-registered meta-research assessment on neuroimaging-based models in the psychiatric literature, quantitatively examining global and regional sampling issues over recent decades, from a view that has been relatively underexplored. A total of 476 studies (n = 118,137) were included in the current assessment. Based on these findings, we built a comprehensive 5-star rating system to quantitatively evaluate the quality of existing machine learning models for psychiatric diagnoses.

RESULTS

A global sampling inequality in these models was revealed quantitatively (sampling Gini coefficient (G) = 0.81, p < .01), varying across different countries (regions) (e.g., China, G = 0.47; the USA, G = 0.58; Germany, G = 0.78; the UK, G = 0.87). Furthermore, the severity of this sampling inequality was significantly predicted by national economic levels (β =  - 2.75, p < .001, R²_adj = 0.40; r =  - .84, 95% CI: - .41 to - .97), and was plausibly predictable for model performance, with higher sampling inequality for reporting higher classification accuracy. Further analyses showed that lack of independent testing (84.24% of models, 95% CI: 81.0-87.5%), improper cross-validation (51.68% of models, 95% CI: 47.2-56.2%), and poor technical transparency (87.8% of models, 95% CI: 84.9-90.8%)/availability (80.88% of models, 95% CI: 77.3-84.4%) are prevailing in current diagnostic classifiers despite improvements over time. Relating to these observations, model performances were found decreased in studies with independent cross-country sampling validations (all p < .001, BF₁₀ > 15). In light of this, we proposed a purpose-built quantitative assessment checklist, which demonstrated that the overall ratings of these models increased by publication year but were negatively associated with model performance.

CONCLUSIONS

Together, improving sampling economic equality and hence the quality of machine learning models may be a crucial facet to plausibly translating neuroimaging-based diagnostic classifiers into clinical practice.

Insulin and Disorders of Behavioural Flexibility

Behavioural inflexibility is a symptom of neuropsychiatric and neurodegenerative disorders such as Obsessive-Compulsive Disorder, Autism Spectrum Disorder and Alzheimer's Disease, encompassing the maintenance of a behaviour even when no longer appropriate. Recent evidence suggests that insulin signalling has roles apart from its regulation of peripheral metabolism and mediates behaviourally-relevant central nervous system (CNS) functions including behavioural flexibility. Indeed, insulin resistance is reported to generate anxious, perseverative phenotypes in animal models, with the Type 2 diabetes medication metformin proving to be beneficial for disorders including Alzheimer's Disease. Structural and functional neuroimaging studies of Type 2 diabetes patients have highlighted aberrant connectivity in regions governing salience detection, attention, inhibition and memory. As currently available therapeutic strategies feature high rates of resistance, there is an urgent need to better understand the complex aetiology of behaviour and develop improved therapeutics. In this review, we explore the circuitry underlying behavioural flexibility, changes in Type 2 diabetes, the role of insulin in CNS outcomes and mechanisms of insulin involvement across disorders of behavioural inflexibility.

Detecting type 2 diabetes mellitus cognitive impairment using whole-brain functional connectivity

Type 2 diabetes mellitus (T2DM) is closely linked to cognitive decline and alterations in brain structure and function. Resting-state functional magnetic resonance imaging (rs-fMRI) is used to diagnose neurodegenerative diseases, such as cognitive impairment (CI), Alzheimer's disease (AD), and vascular dementia (VaD). However, whether the functional connectivity (FC) of patients with T2DM and mild cognitive impairment (T2DM-MCI) is conducive to early diagnosis remains unclear. To answer this question, we analyzed the rs-fMRI data of 37 patients with T2DM and mild cognitive impairment (T2DM-MCI), 93 patients with T2DM but no cognitive impairment (T2DM-NCI), and 69 normal controls (NC). We achieved an accuracy of 87.91% in T2DM-MCI versus T2DM-NCI classification and 80% in T2DM-NCI versus NC classification using the XGBoost model. The thalamus, angular, caudate nucleus, and paracentral lobule contributed most to the classification outcome. Our findings provide valuable knowledge to classify and predict T2DM-related CI, can help with early clinical diagnosis of T2DM-MCI, and provide a basis for future studies.

Anatomical distance affects functional connectivity at rest in medicine-free obsessive-compulsive disorder

BACKGROUND

Brain functional abnormalities at rest have been observed in obsessive-compulsive disorder (OCD). However, whether and how anatomical distance influences functional connectivity (FC) at rest is ambiguous in OCD.

METHODS

Using resting-state functional magnetic resonance imaging data, we calculated the FC of each voxel in the whole-brain and divided FC into short- and long-range FCs in 40 medicine-free patients with OCD and 40 healthy controls (HCs). A support vector machine (SVM) was used to determine whether the altered short- and long-range FCs could be utilized to distinguish OCD from HCs.

RESULTS

Patients had lower short-range positive FC (spFC) and long-range positive FC (lpFC) in the left precentral/postcentral gyrus (t = -5.57 and -5.43; P < 0.05, GRF corrected) and higher lpFC in the right thalamus/caudate, left thalamus, left inferior parietal lobule (IPL) and left cerebellum CrusI/VI (t = 4.59, 4.61, 4.41, and 5.93; P < 0.05, GRF corrected). Furthermore, lower spFC in the left precentral/postcentral gyrus might be used to distinguish OCD from HCs with an accuracy of 80.77%, a specificity of 81.58%, and a sensitivity of 80.00%.

CONCLUSION

These findings highlight that anatomical distance has an effect on the whole-brain FC patterns at rest in OCD. Meanwhile, lower spFC in the left precentral/postcentral gyrus might be applied in distinguishing OCD from HCs.

Autoantibody-associated psychiatric symptoms and syndromes in adults: A narrative review and proposed diagnostic approach

Background

Autoimmune-mediated encephalitis is a disease that often encompasses psychiatric symptoms as its first clinical manifestation’s predominant and isolated characteristic. Novel guidelines even distinguish autoimmune psychosis from autoimmune encephalitis. The aim of this review is thus to explore whether a wide range of psychiatric symptoms and syndromes are associated or correlate with autoantibodies.

Methods

We conducted a PubMed search to identify appropriate articles concerning serum and/or cerebrospinal fluid (CSF) autoantibodies associated with psychiatric symptoms and syndromes between 2000 and 2020. Relying on this data, we developed a diagnostic approach to optimize the detection of autoantibodies in psychiatric patients, potentially leading to the approval of an immunotherapy.

Results

We detected 10 major psychiatric symptoms and syndromes often reported to be associated with serum and/or CSF autoantibodies comprising altered consciousness, disorientation, memory impairment, obsessive-compulsive behavior, psychosis, catatonia, mood dysfunction, anxiety, behavioral abnormalities (autism, hyperkinetic), and sleeping dysfunction. The following psychiatric diagnoses were associated with serum and/or CSF autoantibodies: psychosis and schizophrenia spectrum disorders, mood disorders, minor and major neurocognitive impairment, obsessive-compulsive disorder, autism spectrum disorders (ASD), attention deficit hyperactivity disorder (ADHD), anxiety disorders, eating disorders and addiction. By relying on these symptom clusters and diagnoses in terms of onset and their duration, we classified a subacute or subchronic psychiatric syndrome in patients that should be screened for autoantibodies. We propose further diagnostics entailing CSF analysis, electroencephalography and magnetic resonance imaging of the brain. Exploiting these technologies enables standardized and accurate diagnosis of autoantibody-associated psychiatric symptoms and syndromes to deliver early immunotherapy.

Conclusions

We have developed a clinical diagnostic pathway for classifying subgroups of psychiatric patients whose psychiatric symptoms indicate a suspected autoimmune origin.

•

Autoantibodies are associated with a broad spectrum of psychiatric syndromes.

•

More systematic studies are needed to elucidate the significance of autoantibodies.

•

We developed a pathway to identify autoantibody-associated psychiatric syndromes.

Gray Matter Abnormalities of Orbitofrontal Cortex and Striatum in Drug-Naïve Adult Patients With Obsessive-Compulsive Disorder

Objective: This study examined whether obsessive-compulsive disorder (OCD) patients have gray matter abnormalities in regions related to executive function, and whether such abnormalities are associated with impaired executive function. Methods: Multiple scales were administered to 27 first-episode drug-naïve OCD patients and 29 healthy controls. Comprehensive brain morphometric indicators of orbitofrontal cortex (OFC) and three striatum areas (caudate, putamen, and pallidum) were determined. Hemisphere lateralization index was calculated for each region of interest. Correlations between lateralization index and psychological variables were examined in OCD group. Results: The OCD group had greater local gyrification index for the right OFC and greater gray matter volumes of the bilateral putamen and left pallidum than healthy controls. They also had weaker left hemisphere superiority for local gyrification index of the OFC and gray matter volume of the putamen, but stronger left hemisphere superiority for gray matter volume of the pallidum. Patients' lateralization index for local gyrification index of the OFC correlated negatively with Yale-Brown Obsessive Compulsive Scale and Dysexecutive Questionnaire scores, respectively. Conclusion: Structural abnormalities of the bilateral putamen, left pallidum, and right OFC may underlie OCD pathology. Abnormal lateralization in OCD may contribute to the onset of obsessive-compulsive symptoms and impaired executive function.

Computational Predictions for OCD Pathophysiology and Treatment: A Review

Obsessive compulsive disorder (OCD) can manifest as a debilitating disease with high degrees of co-morbidity as well as clinical and etiological heterogenity. However, the underlying pathophysiology is not clearly understood. Computational psychiatry is an emerging field in which behavior and its neural correlates are quantitatively analyzed and computational models are developed to improve understanding of disorders by comparing model predictions to observations. The aim is to more precisely understand psychiatric illnesses. Such computational and theoretical approaches may also enable more personalized treatments. Yet, these methodological approaches are not self-evident for clinicians with a traditional medical background. In this mini-review, we summarize a selection of computational OCD models and computational analysis frameworks, while also considering the model predictions from a perspective of possible personalized treatment. The reviewed computational approaches used dynamical systems frameworks or machine learning methods for modeling, analyzing and classifying patient data. Bayesian interpretations of probability for model selection were also included. The computational dissection of the underlying pathology is expected to narrow the explanatory gap between the phenomenological nosology and the neuropathophysiological background of this heterogeneous disorder. It may also contribute to develop biologically grounded and more informed dimensional taxonomies of psychopathology.

Chinese College Students Have Higher Anxiety in New Semester of Online Learning During COVID-19: A Machine Learning Approach

The COVID-19 pandemic has caused tremendous loss starting from early this year. This article aims to investigate the change of anxiety severity and prevalence among non-graduating undergraduate students in the new semester of online learning during COVID-19 in China and also to evaluate a machine learning model based on the XGBoost model. A total of 1172 non-graduating undergraduate students aged between 18 and 22 from 34 provincial-level administrative units and 260 cities in China were enrolled onto this study and asked to fill in a sociodemographic questionnaire and the Self-Rating Anxiety Scale (SAS) twice, respectively, during February 15 to 17, 2020, before the new semester started, and March 15 to 17, 2020, 1 month after the new semester based on online learning had started. SPSS 22.0 was used to conduct t-test and single factor analysis. XGBoost models were implemented to predict the anxiety level of students 1 month after the start of the new semester. There were 184 (15.7%, Mean = 58.45, SD = 7.81) and 221 (18.86%, Mean = 57.68, SD = 7.58) students who met the cut-off of 50 and were screened as positive for anxiety, respectively, in the two investigations. The mean SAS scores in the second test was significantly higher than those in the first test (P < 0.05). Significant differences were also found among all males, females, and students majoring in arts and sciences between the two studies (P < 0.05). The results also showed students from Hubei province, where most cases of COVID-19 were confirmed, had a higher percentage of participants meeting the cut-off of being anxious. This article applied machine learning to establish XGBoost models to successfully predict the anxiety level and changes of anxiety levels 4 weeks later based on the SAS scores of the students in the first test. It was concluded that, during COVID-19, Chinese non-graduating undergraduate students showed higher anxiety in the new semester based on online learning than before the new semester started. More students from Hubei province had a different level of anxiety than other provinces. Families, universities, and society as a whole should pay attention to the psychological health of non-graduating undergraduate students and take measures accordingly. It also confirmed that the XGBoost model had better prediction accuracy compared to the traditional multiple stepwise regression model on the anxiety status of university students.

Modeling essential connections in obsessive-compulsive disorder patients using functional MRI

OBJECT

Obsessive-compulsive disorder (OCD) is a mental disease in which people experience uncontrollable and repetitive thoughts or behaviors. Clinical diagnosis of OCD is achieved by using neuropsychological assessment metrics, which are often subjectively affected by psychologists and patients. In this study, we propose a classification model for OCD diagnosis using functional MR images.

METHODS

Using functional connectivity (FC) matrices calculated from brain region of interest (ROI) pairs, a novel Riemann Kernel principal component analysis (PCA) model is employed for feature extraction, which preserves the topological information in the FC matrices. Hierarchical features are then fed into an ensemble classifier based on the XGBoost algorithm. Finally, decisive features extracted during classification are used to investigate the brain FC variations between patients with OCD and healthy controls.

RESULTS

The proposed algorithm yielded a classification accuracy of 91.8%. Additionally, the well-known cortico-striatal-thalamic-cortical (CSTC) circuit and cerebellum were found as highly related regions with OCD. To further analyze the cerebellar-related function in OCD, we demarcated cerebellum into three subregions according to their anatomical and functional property. Using these three functional cerebellum regions as seeds for brain connectivity computation, statistical results showed that patients with OCD have decreased posterior cerebellar connections.

CONCLUSIONS

This study provides a new and efficient method to characterize patients with OCD using resting-state functional MRI. We also provide a new perspective to analyze disease-related features. Despite of CSTC circuit, our model-driven feature analysis reported cerebellum as an OCD-related region. This paper may provide novel insight to the understanding of genetic etiology of OCD.