Jennifer W. Evans’s research while affiliated with National Institute of Mental Health, National Institutes of Health and other places

Visualizations are vital for communicating scientific results. Historically, neuroimaging figures have only depicted regions that surpass a given statistical threshold. This practice substantially biases interpretation of the results and subsequent meta-analyses, particularly towards non-reproducibility. Here we advocate for a "transparent thresholding" approach that not only highlights statistically significant regions but also includes subthreshold locations, which provide key experimental context. This balances the dual needs of distilling modeling results and enabling informed interpretations for modern neuroimaging. We present four examples that demonstrate the many benefits of transparent thresholding, including: removing ambiguity, decreasing hypersensitivity to non-physiological features, catching potential artifacts, improving cross-study comparisons, reducing non-reproducibility biases, and clarifying interpretations. We also demonstrate the many software packages that implement transparent thresholding, several of which were added or streamlined recently as part of this work. A point-counterpoint discussion addresses issues with thresholding raised in real conversations with researchers in the field. We hope that by showing how transparent thresholding can drastically improve the interpretation (and reproducibility) of neuroimaging findings, more researchers will adopt this method.

Visualizations are vital for communicating scientific results. Historically, neuroimaging figures have only depicted regions that surpass a given statistical threshold. This practice substantially biases interpretation of the results and subsequent meta-analyses, particularly towards non-reproducibility. Here we advocate for a “transparent thresholding” approach that not only highlights statistically significant regions but also includes subthreshold locations, which provide key experimental context. This balances the dual needs of distilling modeling results and enabling informed interpretations for modern neuroimaging. We present four examples that demonstrate the many benefits of transparent thresholding, including: removing ambiguity, decreasing hypersensitivity to non-physiological features, catching potential artifacts, improving cross-study comparisons, reducing non-reproducibility biases, and clarifying interpretations. We also demonstrate the many software packages that implement transparent thresholding, several of which were added or streamlined recently as part of this work. A point-counterpoint discussion addresses issues with thresholding raised in real conversations with researchers in the field. We hope that by showing how transparent thresholding can drastically improve the interpretation (and reproducibility) of neuroimaging findings, more researchers will adopt this method.

Importance: Major depressive disorder (MDD) is highly heterogeneous, with marked individual differences in clinical presentation and neurobiology, which may obscure identification of structural brain abnormalities in MDD. To explore this, we used normative modeling to index regional patterns of variability in cortical thickness (CT) across individual patients. Objective: To use normative modeling in a large dataset from the ENIGMA MDD consortium to obtain individualised CT deviations from the norm (relative to age, sex and site) and examine the relationship between these deviations and clinical characteristics. Design, setting, and participants: A normative model adjusting for age, sex and site effects was trained on 35 CT measures from FreeSurfer parcellation of 3,181 healthy controls (HC) from 34 sites (40 scanners). Individualised z-score deviations from this norm for each CT measure were calculated for a test set of 2,119 HC and 3,645 individuals with MDD. For each individual, each CT z-score was classified as being within the normal range (95% of individuals) or within the extreme range (2.5% of individuals with the thinnest or thickest cortices). Main outcome measures: Z-score deviations of CT measures of MDD individuals as estimated from a normative model based on HC. Results: Z-score distributions of CT measures were largely overlapping between MDD and HC (minimum 92%, range 92-98%), with overall thinner cortices in MDD. 34.5% of MDD individuals, and 30% of HC individuals, showed an extreme deviation in at least one region, and these deviations were widely distributed across the brain. There was high heterogeneity in the spatial location of CT deviations across individuals with MDD: a maximum of 12% of individuals with MDD showed an extreme deviation in the same location. Extreme negative CT deviations were associated with having an earlier onset of depression and more severe depressive symptoms in the MDD group, and with higher BMI across MDD and HC groups. Extreme positive deviations were associated with being remitted, of not taking antidepressants and less severe symptoms. Conclusions and relevance: Our study illustrates a large heterogeneity in the spatial location of CT abnormalities across patients with MDD and confirms a substantial overlap of CT measures with HC. We also demonstrate that individualised extreme deviations can identify protective factors and individuals with a more severe clinical picture.

Major depressive disorder (MDD) is a complex psychiatric disorder that affects the lives of hundreds of millions of individuals around the globe. Even today, researchers debate if morphological alterations in the brain are linked to MDD, likely due to the heterogeneity of this disorder. The application of deep learning tools to neuroimaging data, capable of capturing complex non-linear patterns, has the potential to provide diagnostic and predictive biomarkers for MDD. However, previous attempts to demarcate MDD patients and healthy controls (HC) based on segmented cortical features via linear machine learning approaches have reported low accuracies. In this study, we used globally representative data from the ENIGMA-MDD working group containing 7,012 participants from 30 sites (N=2,772 MDD and N=4,240 HC), which allows a comprehensive analysis with generalizable results. Based on the hypothesis that integration of vertex-wise cortical features can improve classification performance, we evaluated the classification of a DenseNet and a Support Vector Machine (SVM), with the expectation that the former would outperform the latter. As we analyzed a multi-site sample, we additionally applied the ComBat harmonization tool to remove potential nuisance effects of site. We found that both classifiers exhibited close to chance performance (balanced accuracy DenseNet: 51%; SVM: 53%), when estimated on unseen sites. Slightly higher classification performance (balanced accuracy DenseNet: 58%; SVM: 55%) was found when the cross-validation folds contained subjects from all sites, indicating site effect. In conclusion, the integration of vertex-wise morphometric features and the use of the non-linear classifier did not lead to the differentiability between MDD and HC. Our results support the notion that MDD classification on this combination of features and classifiers is unfeasible. Future studies are needed to determine whether more sophisticated integration of information from other MRI modalities such as fMRI and DWI will lead to a higher performance in this diagnostic task.

Previous studies have suggested that alterations in white matter (WM) microstructure are implicated in suicidal thoughts and behaviours (STBs). However, findings of diffusion tensor imaging (DTI) studies have been inconsistent. In this large-scale mega-analysis conducted by the ENIGMA Suicidal Thoughts and Behaviours (ENIGMA-STB) consortium, we examined WM alterations associated with STBs. Data processing was standardised across sites, and resulting WM microstructure measures (fractional anisotropy, axial diffusivity, mean diffusivity and radial diffusivity) for 25 WM tracts were pooled across 40 cohorts. We compared these measures among individuals with a psychiatric diagnosis and lifetime history of suicide attempt (n=652; mean age=35.4, sd=14.7; female=71.8%), individuals with a psychiatric diagnosis but no STB (i.e., clinical controls; n=1871; mean age=34, sd=14.8; female=59.8%), and individuals with no mental disorder diagnosis and no STB (i.e., healthy controls; n=642; mean age=29.6, sd=13.1; female=62.9%). We also compared these measures among individuals with recent suicidal ideation (n=714; mean age=36.3, sd=15.3; female=66.1%), clinical controls (n=1184; mean age=36.8, sd=15.6; female=63.1%), and healthy controls (n=1240; mean age= 31.6, sd=15.5; female=61.0%). We found subtle but statistically significant effects, such as lower fractional anisotropy associated with a history of suicide attempt, over and above the effect of psychiatric diagnoses. These effects were strongest in the corona radiata, thalamic radiation, fornix/stria terminalis, corpus callosum and superior longitudinal fasciculus. Effect sizes were small (Cohens d < 0.25). Recent suicidal ideation was not associated with alterations in WM microstructure. This large-scale coordinated mega-analysis revealed subtle regional and global alterations in WM microstructure in individuals with a history of suicide attempt. Longitudinal studies are needed to confirm whether these alterations are a risk factor for suicidal behaviour.

The hippocampus and amygdala have been implicated in the pathophysiology and treatment of major depressive disorder (MDD). Preclinical models suggest that stress-related changes in these regions can be reversed by antidepressants, including ketamine. Clinical studies have identified reduced volumes in MDD that are thought to be potentiated by early life stress and worsened by repeated depressive episodes. This study used 3T and 7T structural magnetic resonance imaging data to examine longitudinal changes in hippocampal and amygdalar subfield volumes associated with ketamine treatment. Data were drawn from a previous double-blind, placebo-controlled, crossover trial of healthy volunteers (HVs) unmedicated individuals with treatment-resistant depression (TRD) (3T: 18 HV, 26 TRD, 7T: 17 HV, 30 TRD) who were scanned at baseline and twice following either a 40 min IV ketamine (0.5 mg/kg) or saline infusion (acute: 1–2 days, interim: 9–10 days post infusion). No baseline differences were noted between the two groups. At 10 days post-infusion, a slight increase was observed between ketamine and placebo scans in whole left amygdalar volume in individuals with TRD. No other differences were found between individuals with TRD and HVs at either field strength. These findings shed light on the timing of ketamine’s effects on cortical structures.

Machine learning (ML) techniques have gained popularity in the neuroimaging field due to their potential for classifying neuropsychiatric disorders. However, the diagnostic predictive power of the existing algorithms has been limited by small sample sizes, lack of representativeness, data leakage, and/or overfitting. Here, we overcome these limitations with the largest multi-site sample size to date (N = 5365) to provide a generalizable ML classification benchmark of major depressive disorder (MDD) using shallow linear and non-linear models. Leveraging brain measures from standardized ENIGMA analysis pipelines in FreeSurfer, we were able to classify MDD versus healthy controls (HC) with a balanced accuracy of around 62%. But after harmonizing the data, e.g., using ComBat, the balanced accuracy dropped to approximately 52%. Accuracy results close to random chance levels were also observed in stratified groups according to age of onset, antidepressant use, number of episodes and sex. Future studies incorporating higher dimensional brain imaging/phenotype features, and/or using more advanced machine and deep learning methods may yield more encouraging prospects.