Identifying neuroanatomical signatures of anorexia nervosa: a multivariate machine learning approach

Review of machine learning solutions for eating disorders

BACKGROUND

Eating Disorders (EDs) are one of the most complex psychiatric disorders, with significant impairment of psychological and physical health, and psychosocial functioning, and are associated with low rates of early detection, low recovery and high relapse rates. This underscores the need for better diagnostic and treatment methods.

OBJECTIVE

This narrative review explores current Machine Learning (ML) and Artificial Intelligence (AI) applications in the domain of EDs, with a specific emphasis on clinical management in treatment settings. The primary objective are to (i) decrease the knowledge gap between ED researchers and AI-practitioners, by presenting the current state-of-the-art AI applications (including models for causality) in different ED use-cases; (ii) identify limitations of these existing AI interventions and how to address them.

RESULTS

AI/ML methods have been applied in different ED use-cases, including ED risk factor identification and incidence prediction (including the analysis of social media content in the general population), diagnosis, monitoring patients and treatment response and prognosis in clinical populations. A comparative analysis of AI-techniques deployed in these use-cases have been performed, considering factors such as complexity, flexibility, functionality, explainability and adaptability to healthcare constraints.

CONCLUSION

Multiple restrictions have been identified in the existing methods in ML and Causality in terms of achieving actionable healthcare for ED, like lack of good quality and quantity of data for models to train on, while requiring models to be flexible, high-performing, yet being explainable and producing counterfactual explanations, for ensuring the fairness and trustworthiness of its decisions. We conclude that to overcome these limitations and for future AI research and application in clinical management of ED, (i) careful considerations are required with regards to AI-model selection, and (ii) joint efforts from ED researcher and patient community are essential in building better quality and quantity of dedicated ED datasets and secure AI-solution framework.

Brain ventricle and choroid plexus morphology as predictor of treatment response in major depression: Findings from the EMBARC study

Recent observations suggest a role of the volume of the cerebral ventricle volume, corpus callosum (CC) segment volume, in particular that of the central-anterior part, and choroid plexus (CP) volume for treatment resistance of major depressive disorder (MDD). An increased CP volume has been associated with increased inflammatory activity and changes in the structure of the ventricles and corpus callosum. We attempt to replicate and confirm that these imaging markers are associated with clinical outcome in subjects from the EMBARC study, as implied by a recent pilot study. The EMBARC study is a placebo controlled randomized study comparing sertraline vs. placebo in patients with MDD to identify biological markers of therapy resistance. Association of baseline volumes of the lateral ventricles (LVV), choroid plexus volume (CPV) and volume of segments of the CC with treatment response after 4 weeks treatment was evaluated. 171 subjects (61 male, 110 female) completed the 4 week assessments; gender and age were taken into account for this analyses. As previously reported, no treatment effect of sertraline vs. placebo was observed, therefore the study characterized prognostic markers of response in the pooled population. Change in depression severity was identified by the ratio of the Hamilton-Depression rating scale 17 (HAMD-17) at week 4 divided by the HAMD-17 at baseline (HAMD-17 ratio). Volumes of the lateral ventricles and choroid plexi were positively correlated with the HAMD-17 ratio, indication worse outcome with larger ventricles and choroid plexus volumes, whereas the volume of the central-anterior corpus callosum was negatively correlated with the HAMD-17 ratio. Responders (n = 54) had significantly smaller volumes of the lateral ventricles and CP compared to non-responders (n = 117), whereas the volume of mid-anterior CC was significantly larger compared to non-responders (n = 117), confirming our previous findings. In an exploratory way associations between enlarged LVV and CPV and signs of lipid dysregulation were observed. In conclusion, we confirmed that volumes of lateral ventricles, choroid plexi and the mid-anterior corpus callosum are associated with clinical improvement of depression and may be indicators of metabolic/inflammatory activity.

Machine learning research based on diffusion tensor images to distinguish between anorexia nervosa and bulimia nervosa

Background

Anorexia nervosa (AN) and bulimia nervosa (BN), two subtypes of eating disorders, often present diagnostic challenges due to their overlapping symptoms. Machine learning has proven its capacity to improve group classification without requiring researchers to specify variables. The study aimed to distinguish between AN and BN using machine learning models based on diffusion tensor images (DTI).

Methods

This is a cross-sectional study, drug-naive females diagnosed with anorexia nervosa (AN) and bulimia nervosa (BN) were included. Demographic data and DTI were collected for all patients. Features for machine learning included Fractional anisotropy (FA), axial diffusivity (AD), radial diffusivity (RD), and mean diffusivity (MD). Support vector machine was constructed by LIBSVM, MATLAB2013b, and FSL5.0.9 software.

Results

A total of 58 female patients (24 AN, 34 BN) were included in this study. Statistical analysis revealed no significant differences in age, years of education, or course of illness between the two groups. AN patients had significantly lower BMI than BN patients. The AD model exhibited an area under the curve was 0.793 (accuracy: 75.86%, sensitivity: 66.67%, specificity: 88.23%), highlighting the left middle temporal gyrus (MTG_L) and the left superior temporal gyrus (STG_L) as differentiating brain regions. AN patients exhibited lower AD features in the STG_L and MTG_L than BN. Machine learning analysis indicated no significant differences in FA, MD, and RD values between AN and BN groups (p > 0.001).

Conclusion

Machine learning based on DTI could effectively distinguish between AN and BN, with MTG_L and STG_L potentially serving as neuroimaging biomarkers.

Predicting long-term outcome in anorexia nervosa: a machine learning analysis of brain structure at different stages of weight recovery

BACKGROUND

Anorexia nervosa (AN) is characterized by sizable, widespread gray matter (GM) reductions in the acutely underweight state. However, evidence for persistent alterations after weight-restoration has been surprisingly scarce despite high relapse rates, frequent transitions to other psychiatric disorders, and generally unfavorable outcome. While most studies investigated brain regions separately (univariate analysis), psychiatric disorders can be conceptualized as brain network disorders characterized by multivariate alterations with only subtle local effects. We tested for persistent multivariate structural brain alterations in weight-restored individuals with a history of AN, investigated their putative biological substrate and relation with 1-year treatment outcome.

METHODS

We trained machine learning models on regional GM measures to classify healthy controls (HC) (N = 289) from individuals at three stages of AN: underweight patients starting intensive treatment (N = 165, used as baseline), patients after partial weight-restoration (N = 115), and former patients after stable and full weight-restoration (N = 89). Alterations after weight-restoration were related to treatment outcome and characterized both anatomically and functionally.

RESULTS

Patients could be classified from HC when underweight (ROC-AUC = 0.90) but also after partial weight-restoration (ROC-AUC = 0.64). Alterations after partial weight-restoration were more pronounced in patients with worse outcome and were not detected in long-term weight-recovered individuals, i.e. those with favorable outcome. These alterations were more pronounced in regions with greater functional connectivity, not merely explained by body mass index, and even increases in cortical thickness were observed (insula, lateral orbitofrontal, temporal pole).

CONCLUSIONS

Analyzing persistent multivariate brain structural alterations after weight-restoration might help to develop personalized interventions after discharge from inpatient treatment.

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.

Brain Ventricle and Choroid Plexus Morphology as Predictor of Treatment Response: Findings from the EMBARC Study

Recent observations suggest a role of the choroid plexus (CP) and cerebral ventricle volume (CV), to identify treatment resistance of major depressive disorder (MDD). We tested the hypothesis that these markers are associated with clinical improvement in subjects from the EMBARC study, as implied by a recent pilot study. The EMBARC study characterized biological markers in a randomized placebo-controlled trial of sertraline vs. placebo in patients with MDD. Association of baseline volumes of CV, CP and of the corpus callosum (CC) with treatment response after 4 weeks treatment were evaluated. 171 subjects (61 male, 110 female) completed the 4 week assessments; gender, site and age were taken into account for this analyses. As previously reported, no treatment effect of sertraline was observed, but prognostic markers for clinical improvement were identified. Responders (n = 54) had significantly smaller volumes of the CP and lateral ventricles, whereas the volume of mid-anterior and mid-posterior CC was significantly larger compared to non-responders (n = 117). A positive correlation between CV volume and CP volume was observed, whereas a negative correlation between CV volume and both central-anterior and central-posterior parts of the CC emerged. In an exploratory way correlations between enlarged VV and CP volume on the one hand and signs of metabolic syndrome, in particular triglyceride plasma concentrations, were observed. A primary abnormality of CP function in MDD may be associated with increased ventricles, compression of white matter volume, which may affect treatment response speed or outcome. Metabolic markers may mediate this relationship.

Volume and complexity of the thalamus in Anorexia Nervosa: An exploratory evaluation

OBJECTIVE

Recent neuroscientific findings have highlighted the role of the thalamus in several cognitive functions, ranging from perception to cognitive flexibility, memory, and body representation. Since some of these functions may be involved in the pathophysiology of Anorexia Nervosa (AN), this study aims at exploring thalamic structure in different phases of the disorder.

METHOD

The sample included 38 patients with acute AN, 20 patients who fully recovered from AN (recAN), and 38 healthy controls (HC), all female. All participants underwent high-resolution MRI. The volumes of the whole thalamus and 25 thalamic nuclei were extracted using an automated segmentation algorithm, and thalamic fractal dimension was estimated using the calcFD toolbox.

RESULTS

Patients with acute AN, compared to HC, displayed reduced thalamic volume and complexity both at the whole level and at the level of specific nuclei. In patients recAN, instead, alterations were observed only at the level of the right laterodorsal and central lateral nuclei.

CONCLUSIONS

In the acute phase of the disorder patients with AN present a widespread reduction in thalamic volume and complexity. However, these alterations seem to normalise almost completely following weight restoration, thus suggesting the involvement of malnutrition-related mechanisms.

Evaluation of Risk of Bias in Neuroimaging-Based Artificial Intelligence Models for Psychiatric Diagnosis: A Systematic Review

IMPORTANCE

Neuroimaging-based artificial intelligence (AI) diagnostic models have proliferated in psychiatry. However, their clinical applicability and reporting quality (ie, feasibility) for clinical practice have not been systematically evaluated.

OBJECTIVE

To systematically assess the risk of bias (ROB) and reporting quality of neuroimaging-based AI models for psychiatric diagnosis.

EVIDENCE REVIEW

PubMed was searched for peer-reviewed, full-length articles published between January 1, 1990, and March 16, 2022. Studies aimed at developing or validating neuroimaging-based AI models for clinical diagnosis of psychiatric disorders were included. Reference lists were further searched for suitable original studies. Data extraction followed the CHARMS (Checklist for Critical Appraisal and Data Extraction for Systematic Reviews of Prediction Modeling Studies) and PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-analyses) guidelines. A closed-loop cross-sequential design was used for quality control. The PROBAST (Prediction Model Risk of Bias Assessment Tool) and modified CLEAR (Checklist for Evaluation of Image-Based Artificial Intelligence Reports) benchmarks were used to systematically evaluate ROB and reporting quality.

FINDINGS

A total of 517 studies presenting 555 AI models were included and evaluated. Of these models, 461 (83.1%; 95% CI, 80.0%-86.2%) were rated as having a high overall ROB based on the PROBAST. The ROB was particular high in the analysis domain, including inadequate sample size (398 of 555 models [71.7%; 95% CI, 68.0%-75.6%]), poor model performance examination (with 100% of models lacking calibration examination), and lack of handling data complexity (550 of 555 models [99.1%; 95% CI, 98.3%-99.9%]). None of the AI models was perceived to be applicable to clinical practices. Overall reporting completeness (ie, number of reported items/number of total items) for the AI models was 61.2% (95% CI, 60.6%-61.8%), and the completeness was poorest for the technical assessment domain with 39.9% (95% CI, 38.8%-41.1%).

CONCLUSIONS AND RELEVANCE

This systematic review found that the clinical applicability and feasibility of neuroimaging-based AI models for psychiatric diagnosis were challenged by a high ROB and poor reporting quality. Particularly in the analysis domain, ROB in AI diagnostic models should be addressed before clinical application.

Application of artificial intelligence tools in diagnosis and treatmentof mental disorders

Introduction: Artificial intelligence research is increasing its application in mental health services. Machine learning, deep learning, semantic analysis in the form of transcriptions of patients' statements enable early diagnosis of psychotic disorders, ADHD, anorexia nervosa. Of great importance are the so-called digital therapists. This paper aims to show the use of AI tools in diagnosing, treating, the benefits and limitations associated with mental disorders. Material and methodS: This literature review was conducted by searching scientific articles from 2015 to 2022. The basis were PubMED, OpenKnowledge, Web of Science, using the following keywords: artificial intelligence, digital therapy, psychiatry, machine learning. Results: A review indicates the widespread use of AI tools in screening for mental disorders. These tools advance the clinical diagnosis medical specialists make up for several years. They impact solving medical staff shortages, lack of access to medical facilities and leveling patient resistance to treatment. The benefits are ultra-fast analysis of large sets of information, effective screening of people in need of specialized psychiatric care, reduction of doctors' duties and maximization of their work efficiency. During the current COVID 19 pandemic, robots in the form of digital psychotherapists are playing a special role. Conclusions: The need for further research, testing and clarification of regulations related to the use of AI tools is indicated. Ethical and social problems need to be resolved. The tools should not form the basis of autonomous therapy without the supervision of highly trained professionals. Human beings should be at the center of analysis just as their health and well-being. Keywords: artificial intelligence, digital therapy, psychiatry, machine learning

Examining the utility of nonlinear machine learning approaches versus linear regression for predicting body image outcomes: The U.S. Body Project I

Most body image studies assess only linear relations between predictors and outcome variables, relying on techniques such as multiple Linear Regression. These predictor variables are often validated multi-item measures that aggregate individual items into a single scale. The advent of machine learning has made it possible to apply Nonlinear Regression algorithms-such as Random Forest and Deep Neural Networks-to identify potentially complex linear and nonlinear connections between a multitude of predictors (e.g., all individual items from a scale) and outcome (output) variables. Using a national dataset, we tested the extent to which these techniques allowed us to explain a greater share of the variance in body-image outcomes (adjusted R²) than possible with Linear Regression. We examined how well the connections between body dissatisfaction and dieting behavior could be predicted from demographic factors and measures derived from objectification theory and the tripartite-influence model. In this particular case, although Random Forest analyses sometimes provided greater predictive power than Linear Regression models, the advantages were small. More generally, however, this paper demonstrates how body image researchers might harness the power of machine learning techniques to identify previously undiscovered relations among body image variables.

Potential benefits and limitations of machine learning in the field of eating disorders: current research and future directions

Advances in machine learning and digital data provide vast potential for mental health predictions. However, research using machine learning in the field of eating disorders is just beginning to emerge. This paper provides a narrative review of existing research and explores potential benefits, limitations, and ethical considerations of using machine learning to aid in the detection, prevention, and treatment of eating disorders. Current research primarily uses machine learning to predict eating disorder status from females' responses to validated surveys, social media posts, or neuroimaging data often with relatively high levels of accuracy. This early work provides evidence for the potential of machine learning to improve current eating disorder screening methods. However, the ability of these algorithms to generalise to other samples or be used on a mass scale is only beginning to be explored. One key benefit of machine learning over traditional statistical methods is the ability of machine learning to simultaneously examine large numbers (100s to 1000s) of multimodal predictors and their complex non-linear interactions, but few studies have explored this potential in the field of eating disorders. Machine learning is also being used to develop chatbots to provide psychoeducation and coping skills training around body image and eating disorders, with implications for early intervention. The use of machine learning to personalise treatment options, provide ecological momentary interventions, and aid the work of clinicians is also discussed. Machine learning provides vast potential for the accurate, rapid, and cost-effective detection, prevention, and treatment of eating disorders. More research is needed with large samples of diverse participants to ensure that machine learning models are accurate, unbiased, and generalisable to all people with eating disorders. There are important limitations and ethical considerations with utilising machine learning methods in practice. Thus, rather than a magical solution, machine learning should be seen as an important tool to aid the work of researchers, and eventually clinicians, in the early identification, prevention, and treatment of eating disorders.

Structural brain differences in recovering and weight-recovered adult outpatient women with anorexia nervosa

BACKGROUND

Anorexia nervosa is a complex psychiatric illness that includes severe low body weight with cognitive distortions and altered eating behaviors. Brain structures, including cortical thicknesses in many regions, are reduced in underweight patients who are acutely ill with anorexia nervosa. However, few studies have examined adult outpatients in the process of recovering from anorexia nervosa. Evaluating neurobiological problems at different physiological stages of anorexia nervosa may facilitate our understanding of the recovery process.

METHODS

Magnetic resonance imaging (MRI) images from 37 partially weight-restored women with anorexia nervosa (pwAN), 32 women with a history of anorexia nervosa maintaining weight restoration (wrAN), and 41 healthy control women were analyzed using FreeSurfer. Group differences in brain structure, including cortical thickness, areas, and volumes, were compared using a series of factorial f-tests, including age as a covariate, and correcting for multiple comparisons with the False Discovery Rate method.

RESULTS

The pwAN and wrAN cohorts differed from each other in body mass index, eating disorder symptoms, and social problem solving orientations, but not depression or self-esteem. Relative to the HC cohort, eight cortical thicknesses were thinner for the pwAN cohort; these regions were predominately right-sided and in the cingulate and frontal lobe. One of these regions, the right pars orbitalis, was also thinner for the wrAN cohort. One region, the right parahippocampal gyrus, was thicker in the pwAN cohort. One volume, the right cerebellar white matter, was reduced in the pwAN cohort. There were no differences in global white matter, gray matter, or subcortical volumes across the cohorts.

CONCLUSIONS

Many regional structural differences were observed in the pwAN cohort with minimal differences in the wrAN cohort. These data support a treatment focus on achieving and sustaining full weight restoration to mitigate possible neurobiological sequela of AN. In addition, the regions showing cortical thinning are similar to structural changes reported elsewhere for suicide attempts, anxiety disorders, and autistic spectrum disorder. Understanding how brain structure and function are related to clinical symptoms expressed during the course of recovering from AN is needed.

Current directions in biomarkers and endophenotypes for anorexia nervosa: A scoping review

There are currently no validated biomarkers for anorexia nervosa (AN), though recent literature suggests an increased research interest in this area. Biomarkers are objective, measurable indicators of illness that can be used to assist with diagnosis, risk assessment, and tracking of illness state. Related to biomarkers are endophenotypes, which are quantifiable phenomena that are distinct from symptoms and which link genes to manifest illness. In this scoping review, we sought to provide a summary of recent research conducted in the pursuit of biomarkers and endophenotypes for AN. The findings indicate that a number of possible biomarkers which can assess the presence or severity of AN independently of weight status, including psychophysical (e.g., eye-tracking) and biological (e.g., immune, endocrine, metabolomic, neurobiological) markers, are currently under investigation. However, this research is still in early phases and lacking in replication studies. Endophenotype research has largely been confined to the study of several neurocognitive features, with mixed evidence to support their classification as possible endophenotypes for the disorder. The study of biomarkers and endophenotypes in AN involves significant challenges due to confounding factors of illness-related sequalae, such as starvation. Future research in these areas must prioritise direct evaluation of the sensitivity, specificity and test-retest reliability of proposed biomarkers and enhanced control of confounding physical consequences of AN in the study of biomarkers and endophenotypes.

The song of Anorexia Nervosa: a specific evoked potential response to musical stimuli in affected participants

PURPOSE

Research applying electroencephalography (EEG) to Anorexia Nervosa (AN) is still limited, even though in other psychiatric disorders EEG has permitted to find out the hallmarks of the disorder. The aim of the study was to explore whether EEG basal activity and reactivity to musical stimulation differ in participants with AN as compared to healthy subjects (HS).

METHODS

Twenty female participants (respectively 10 with AN and 10 healthy controls) were administered a battery of psychometric tests and underwent EEG under three different conditions: (1) at baseline; (2) after a generic music stimulation; and (3) after a favorite musical stimulation.

RESULTS

In participants with AN, basal EEG showed the higher absolute amplitude of cortical slow waves (theta) in the parieto-occipital and temporal derivations, with a deficit in the beta band. In AN, there was a higher N100 latency and a reduced P300 latency compared to HS. While the N100 and P300 latencies were sensitive to the musical stimulus in HS, there was no difference after music stimulation in AN.

CONCLUSION

These data suggest that AN is accompanied by a state of brain hyperarousal with abnormal reactivity to environmental stimuli, similar to the state of HS after musical stimulation. If confirmed, this finding may have treatment implications.

LEVEL OF EVIDENCE

III, Evidence obtained from well-designed cohort or case-control analytic studies.

Ventricular volume, white matter alterations and outcome of major depression and their relationship to endocrine parameters - A pilot study

OBJECTIVES

Brain morphology and its relation to endocrine parameters were examined, in order to determine the link of these parameters to treatment outcome to psychopharmacological treatment in depressed patients.

METHODS

We examined the potentially predictive value of Magnetic Resonance Imaging (MRI) parameters related to mineralocorticoid receptor (MR) function on the treatment outcome of depression. 16 inpatients with a major depressive episode (MDE) were studied at baseline and 14 of them approximately six weeks later. Physiological biomarkers and 3-T-structural MRI based volume measures, using FreeSurfer 6.0 software, were determined.

RESULTS

Non-responders (<50% reduction of HAMD-21; n = 6) had a significantly smaller volume of the right anterior cingulate cortex, a significantly larger ventricle to brain ratio (VBR) and third ventricle volume, and smaller volumes of the central and central-anterior corpus callosum (CC) in comparison to responders (n = 7; all p ≤ 0.05). Correlational analysis (Spearman) demonstrated that larger ventricle volume was correlated to a worse treatment outcome, higher body mass index (BMI) and smaller CC segment volume, whereas the total CC volume was negatively correlated to the saliva aldosterone/cortisol concentration ratio (AC-ratio).

CONCLUSION

Large ventricular volume may be a predictive marker for worse treatment response to standard antidepressant treatment, potentially via compression of white matter structures. A mediating role of the previously identified markers BMI and the AC-ratio, is suggested.

Eating Disorder Neuroimaging Initiative (EDNI): a multicentre prospective cohort study protocol for elucidating the neural effects of cognitive-behavioural therapy for eating disorders

INTRODUCTION

Anorexia nervosa is a refractory psychiatric disorder with a mortality rate of 5.9% and standardised mortality ratio of 5.35, which is much higher than other psychiatric disorders. The standardised mortality ratio of bulimia nervosa is 1.49; however, it is characterised by suicidality resulting in a shorter time to death. While there is no current validated drug treatment for eating disorders in Japan, cognitive-behavioural therapy (CBT) is a well-established and commonly used treatment. CBT is also recommended in the Japanese Guidelines for the Treatment of Eating Disorders (2012) and has been covered by insurance since 2018. However, the neural mechanisms responsible for the effect of CBT have not been elucidated, and the use of biomarkers such as neuroimaging data would be beneficial.

METHODS AND ANALYSIS

The Eating Disorder Neuroimaging Initiative is a multisite prospective cohort study. We will longitudinally collect data from 72 patients with eating disorders (anorexia nervosa and bulimia nervosa) and 70 controls. Data will be collected at baseline, after 21-41 sessions of CBT and 12 months later. We will assess longitudinal changes in neural circuit function, clinical data, gene expression and psychological measures by therapeutic intervention and analyse the relationship among them using machine learning methods.

ETHICS AND DISSEMINATION

The study was approved by The Ethical Committee of the National Center of Neurology and Psychiatry (A2019-072). We will obtain written informed consent from all patients who participate in the study after they had been fully informed about the study protocol. All imaging, demographic and clinical data are shared between the participating sites and will be made publicly available in 2024.

TRIAL REGISTRATION NUMBER

UMIN000039841.

Hippocampal volumes in anorexia nervosa at different stages of the disorder

INTRODUCTION

Hippocampus is involved in a range of cognitive and behavioural processes, and its volume has been found to be reduced in different psychiatric disorders. The present study aims at exploring hippocampal volumes in anorexia nervosa (AN) at different stages of the disorder (a few months after onset, more than 1 year after onset and after recovery).

METHODS

Two samples were included in the present study. The first was composed of 58 patients (38 with present AN, 20 full recovered from AN) and 38 age-matched healthy women (HW); the second, recruited at a different site, included 20 patients with AN and 16 HW. Hippocampal volume has been estimated using an automated segmentation algorithm. Age, site of scanning and total intracranial volumes were used as covariates in the statistical analyses.

RESULTS

AN patients showed a reduced hippocampal volume in comparison to HW, with no substantial differences between patients with recent onset and those with a longer duration of illness. Among patients, hippocampal volumes correlated with body mass index, anxiety and drive for thinness.

DISCUSSION

Our findings suggest an early role of malnutrition in the morphologic alterations of the hippocampus in AN and a possible role of this brain structure in mediating specific psychopathological traits.

Multimodal neuroimaging in anorexia nervosa

Anorexia nervosa (AN) is a severe and complex psychiatric disorder characterized by intense fear about weight gain and finalized to food-related control behaviors. Growing interest has been demonstrated about neurobiological processes subtend to AN physiopathology. The present review aimed to collect neurostructural and neurofunctional available data from 2010 to 2019. Results have been organized according to the neuroimaging technique employed, also including a specific section on electroencephalographic results, mostly neglected in previous reviews. Diffuse cerebral vulnerability has been demonstrated and the contribution of several structures has been identified. Insula, cingulate cortex, parietal and frontal areas are primarily involved both by structural and functional perspectives. Moreover, consistent alterations in white matter integrity and brain electrical activity have been reported. Neuroimaging findings give a substantial contribution to AN pathophysiological description, also in order to understand altered but reversible processes in the passage from acute illness phase to disorder's remission, useful also for defining therapy.

Towards Precision Medicine in Psychosis: Benefits and Challenges of Multimodal Multicenter Studies-PSYSCAN: Translating Neuroimaging Findings From Research into Clinical Practice

In the last 2 decades, several neuroimaging studies investigated brain abnormalities associated with the early stages of psychosis in the hope that these could aid the prediction of onset and clinical outcome. Despite advancements in the field, neuroimaging has yet to deliver. This is in part explained by the use of univariate analytical techniques, small samples and lack of statistical power, lack of external validation of potential biomarkers, and lack of integration of nonimaging measures (eg, genetic, clinical, cognitive data). PSYSCAN is an international, longitudinal, multicenter study on the early stages of psychosis which uses machine learning techniques to analyze imaging, clinical, cognitive, and biological data with the aim of facilitating the prediction of psychosis onset and outcome. In this article, we provide an overview of the PSYSCAN protocol and we discuss benefits and methodological challenges of large multicenter studies that employ neuroimaging measures.

Structural neuroimaging as clinical predictor: A review of machine learning applications

In this paper, we provide an extensive overview of machine learning techniques applied to structural magnetic resonance imaging (MRI) data to obtain clinical classifiers. We specifically address practical problems commonly encountered in the literature, with the aim of helping researchers improve the application of these techniques in future works. Additionally, we survey how these algorithms are applied to a wide range of diseases and disorders (e.g. Alzheimer's disease (AD), Parkinson's disease (PD), autism, multiple sclerosis, traumatic brain injury, etc.) in order to provide a comprehensive view of the state of the art in different fields.

Cortical thickness patterns as state biomarker of anorexia nervosa

OBJECTIVE

Only few studies have investigated cortical thickness in anorexia nervosa (AN), and it is unclear whether patterns of altered cortical thickness can be identified as biomarkers for AN.

METHOD

Cortical thickness was measured in 19 adult women with restricting-type AN, 24 individuals recovered from restricting-type AN (REC-AN) and 24 healthy controls. Those individuals with current or recovered from AN had previously shown altered regional cortical volumes across orbitofrontal cortex and insula. A linear relevance vector machine-learning algorithm estimated patterns of regional thickness across 24 subdivisions of those regions.

RESULTS

Region-based analysis showed higher cortical thickness in AN and REC-AN, compared to controls, in the right medial orbital (olfactory) sulcus, and greater cortical thickness for short insular gyri in REC-AN versus controls bilaterally. The machine-learning algorithm identified a pattern of relatively higher right orbital, right insular and left middle frontal cortical thickness, but lower left orbital, right middle and inferior frontal, and bilateral superior frontal cortical thickness specific to AN versus controls (74% specificity and 74% sensitivity, χ² p < .004); predicted probabilities differed significantly between AN and controls (p < .023). No pattern significantly distinguished the REC-AN group from controls.

CONCLUSIONS

Higher cortical thickness in medial orbitofrontal cortex and insula probably contributes to higher gray matter volume in AN in those regions. The machine-learning algorithm identified a mixed pattern of mostly higher orbital and insular, but relatively lower superior frontal cortical thickness in individuals with current AN. These novel results suggest that regional cortical thickness patterns could be state markers for AN.

Associations between cortical thickness, structural connectivity and severity of dimensional bulimia nervosa symptomatology

Bulimia nervosa (BN) is a psychiatric illness defined by preoccupation with body image (cognitive 'symptoms'), binge eating and compensatory behaviors. Although diagnosed BN has been related to grey matter alterations, characterization of brain structure in women with a range of BN symptoms has not been made. This study examined whether cortical thickness (CT) values scaled with severity of BN cognitions in 33 women with variable BN pathology. We then assessed global structural connectivity (SC) of CT to determine if individual differences in global SC relate to BN symptom severity. We used the Eating Disorder Examination Questionnaire (EDE-Q) as a continuous measure of BN symptom severity. EDE-Q score was negatively related to global CT and local CT in the left middle frontal gyrus, right superior frontal gyrus and bilateral orbitofrontal cortex (OFC) and temporoparietal regions. Moreover, cortical thinning was most pronounced in regions with high global connectivity. Finally, individual contributions to global SC at the group level related to EDE-Q score, where increased EDE-Q score correlated with reduced connectivity of the left OFC and middle temporal cortex and increased connectivity of the right superior parietal lobule. Findings represent the first evidence of cortical thinning that relates to cognitive BN symptoms.

Machine Learning Approaches for Clinical Psychology and Psychiatry

Machine learning approaches for clinical psychology and psychiatry explicitly focus on learning statistical functions from multidimensional data sets to make generalizable predictions about individuals. The goal of this review is to provide an accessible understanding of why this approach is important for future practice given its potential to augment decisions associated with the diagnosis, prognosis, and treatment of people suffering from mental illness using clinical and biological data. To this end, the limitations of current statistical paradigms in mental health research are critiqued, and an introduction is provided to critical machine learning methods used in clinical studies. A selective literature review is then presented aiming to reinforce the usefulness of machine learning methods and provide evidence of their potential. In the context of promising initial results, the current limitations of machine learning approaches are addressed, and considerations for future clinical translation are outlined.

Preserved white matter microstructure in young patients with anorexia nervosa?

A massive but reversible reduction of cortical thickness and subcortical gray matter (GM) volumes in Anorexia Nervosa (AN) has been recently reported. However, the literature on alterations in white matter (WM) volume and microstructure changes in both acutely underweight AN (acAN) and after recovery (recAN) is sparse and results are inconclusive. Here, T1-weighted and diffusion-weighted MRI data in a sizable sample of young and medication-free acAN (n = 35), recAN (n = 32), and age-matched female healthy controls (HC, n = 62) were obtained. For analysis, a well-validated global probabilistic tractography reconstruction algorithm including rigorous motion correction implemented in FreeSurfer: TRACULA (TRActs Constrained by UnderLying Anatomy) were used. Additionally, a clustering algorithm and a multivariate pattern classification technique to WM metrics to predict group membership were applied. No group differences in either WM volume or WM microstructure were detected with standard analysis procedures either in acAN or recAN relative to HC after controlling for the number of performed statistical tests. These findings were not affected by age, IQ, or psychiatric symptoms. While cluster analysis was unsuccessful at discriminating between groups, multivariate pattern classification showed some ability to separate acAN from HC (but not recAN from HC). However, these results were not compatible with a straightforward hypothesis of impaired WM microstructure. The current findings suggest that WM integrity is largely preserved in non-chronic AN. This finding stands in contrast to findings in GM, but may help to explain the relatively intact cognitive performance of young patients with AN and provide the basis for the fast recovery of GM structures. Hum Brain Mapp 37:4069-4083, 2016. © 2016 Wiley Periodicals, Inc.

Neural Predictors of Initiating Alcohol Use During Adolescence

OBJECTIVE

Underage drinking is widely recognized as a leading public health and social problem for adolescents in the United States. Being able to identify at-risk adolescents before they initiate heavy alcohol use could have important clinical and public health implications; however, few investigations have explored individual-level precursors of adolescent substance use. This prospective investigation used machine learning with demographic, neurocognitive, and neuroimaging data in substance-naive adolescents to identify predictors of alcohol use initiation by age 18.

METHOD

Participants (N=137) were healthy substance-naive adolescents (ages 12-14) who underwent neuropsychological testing and structural and functional magnetic resonance imaging (sMRI and fMRI), and then were followed annually. By age 18, 70 youths (51%) initiated moderate to heavy alcohol use, and 67 remained nonusers. Random forest classification models identified the most important predictors of alcohol use from a large set of demographic, neuropsychological, sMRI, and fMRI variables.

RESULTS

Random forest models identified 34 predictors contributing to alcohol use by age 18, including several demographic and behavioral factors (being male, higher socioeconomic status, early dating, more externalizing behaviors, positive alcohol expectancies), worse executive functioning, and thinner cortices and less brain activation in diffusely distributed regions of the brain.

CONCLUSIONS

Incorporating a mix of demographic, behavioral, neuropsychological, and neuroimaging data may be the best strategy for identifying youths at risk for initiating alcohol use during adolescence. The identified risk factors will be useful for alcohol prevention efforts and in research to address brain mechanisms that may contribute to early drinking.

Identification and individualized prediction of clinical phenotypes in bipolar disorders using neurocognitive data, neuroimaging scans and machine learning

Diagnosis, clinical management and research of psychiatric disorders remain subjective - largely guided by historically developed categories which may not effectively capture underlying pathophysiological mechanisms of dysfunction. Here, we report a novel approach of identifying and validating distinct and biologically meaningful clinical phenotypes of bipolar disorders using both unsupervised and supervised machine learning techniques. First, neurocognitive data were analyzed using an unsupervised machine learning approach and two distinct clinical phenotypes identified namely; phenotype I and phenotype II. Second, diffusion weighted imaging scans were pre-processed using the tract-based spatial statistics (TBSS) method and 'skeletonized' white matter fractional anisotropy (FA) and mean diffusivity (MD) maps extracted. The 'skeletonized' white matter FA and MD maps were entered into the Elastic Net machine learning algorithm to distinguish individual subjects' phenotypic labels (e.g. phenotype I vs. phenotype II). This calculation was performed to ascertain whether the identified clinical phenotypes were biologically distinct. Original neurocognitive measurements distinguished individual subjects' phenotypic labels with 94% accuracy (sensitivity=92%, specificity=97%). TBSS derived FA and MD measurements predicted individual subjects' phenotypic labels with 76% and 65% accuracy respectively. In addition, individual subjects belonging to phenotypes I and II were distinguished from healthy controls with 57% and 92% accuracy respectively. Neurocognitive task variables identified as most relevant in distinguishing phenotypic labels included; Affective Go/No-Go (AGN), Cambridge Gambling Task (CGT) coupled with inferior fronto-occipital fasciculus and callosal white matter pathways. These results suggest that there may exist two biologically distinct clinical phenotypes in bipolar disorders which can be identified from healthy controls with high accuracy and at an individual subject level. We suggest a strong clinical utility of the proposed approach in defining and validating biologically meaningful and less heterogeneous clinical sub-phenotypes of major psychiatric disorders.

Individualized identification of euthymic bipolar disorder using the Cambridge Neuropsychological Test Automated Battery (CANTAB) and machine learning

BACKGROUND

Previous studies have reported that patients with bipolar disorder (BD) present with cognitive impairments during mood episodes as well as euthymic phase. However, it is still unknown whether reported neurocognitive abnormalities can objectively identify individual BD patients from healthy controls (HC).

METHODS

A total of 21 euthymic BD patients and 21 demographically matched HC were included in the current study. Participants performed the computerized Cambridge Neurocognitive Test Automated Battery (CANTAB) to assess cognitive performance. The least absolute shrinkage selection operator (LASSO) machine learning algorithm was implemented to identify neurocognitive signatures to distinguish individual BD patients from HC.

RESULTS

The LASSO machine learning algorithm identified individual BD patients from HC with an accuracy of 71%, area under receiver operating characteristic curve of 0.7143 and significant at p=0.0053. The LASSO algorithm assigned individual subjects with a probability score (0-healthy, 1-patient). Patients with rapid cycling (RC) were assigned increased probability scores as compared to patients without RC. A multivariate pattern of neurocognitive abnormalities comprising of affective Go/No-go and the Cambridge gambling task was relevant in distinguishing individual patients from HC.

LIMITATIONS

Our study sample was small as we only considered euthymic BD patients and demographically matched HC.

CONCLUSION

Neurocognitive abnormalities can distinguish individual euthymic BD patients from HC with relatively high accuracy. In addition, patients with RC had more cognitive impairments compared to patients without RC. The predictive neurocognitive signature identified in the current study can potentially be used to provide individualized clinical inferences on BD patients.

Utility of Machine-Learning Approaches to Identify Behavioral Markers for Substance Use Disorders: Impulsivity Dimensions as Predictors of Current Cocaine Dependence

BACKGROUND

Identifying objective and accurate markers of cocaine dependence (CD) can innovate its prevention and treatment. Existing evidence suggests that CD is characterized by a wide range of cognitive deficits, most notably by increased impulsivity. Impulsivity is multidimensional and it is unclear which of its various dimensions would have the highest predictive utility for CD. The machine-learning approach is highly promising for discovering predictive markers of disease. Here, we used machine learning to identify multivariate predictive patterns of impulsivity phenotypes that can accurately classify individuals with CD.

METHODS

Current cocaine-dependent users (N = 31) and healthy controls (N = 23) completed the self-report Barratt Impulsiveness Scale-11 and five neurocognitive tasks indexing different dimensions of impulsivity: (1) Immediate Memory Task (IMT), (2) Stop-Signal Task, (3) Delay-Discounting Task (DDT), (4) Iowa Gambling Task (IGT), and (5) Probabilistic Reversal-Learning task. We applied a machine-learning algorithm to all impulsivity measures.

RESULTS

Machine learning accurately classified individuals with CD and predictions were generalizable to new samples (area under the curve of the receiver-operating characteristic curve was 0.912 in the test set). CD membership was predicted by higher scores on motor and non-planning trait impulsivity, poor response inhibition, and discriminability on the IMT, higher delay discounting on the DDT, and poor decision making on the IGT.

CONCLUSION

Our results suggest that multivariate behavioral impulsivity phenotypes can predict CD with high degree of accuracy, which can potentially be used to assess individuals' vulnerability to CD in clinical settings.