Accuracy of automated classification of major depressive disorder as a function of symptom severity.

BACKGROUND: Growing evidence documents the potential of machine learning for developing brain based diagnostic methods for major depressive disorder (MDD). As symptom severity may influence brain activity, we investigated whether the severity of MDD affected the accuracies of machine learned MDD-vs-Control diagnostic classifiers. METHODS: Forty-five medication-free patients with DSM-IV defined MDD and 19 healthy controls participated in the study. Based on depression severity as determined by the Hamilton Rating Scale for Depression (HRSD), MDD patients were sorted into three groups: mild to moderate depression (HRSD 14-19), severe depression (HRSD 20-23), and very severe depression (HRSD ≥ 24). We collected functional magnetic resonance imaging (fMRI) data during both resting-state and an emotional-face matching task. Patients in each of the three severity groups were compared against controls in separate analyses, using either the resting-state or task-based fMRI data. We use each of these six datasets with linear support vector machine (SVM) binary classifiers for identifying individuals as patients or controls. RESULTS: The resting-state fMRI data showed statistically significant classification accuracy only for the very severe depression group (accuracy 66%, p = 0.012 corrected), while mild to moderate (accuracy 58%, p = 1.0 corrected) and severe depression (accuracy 52%, p = 1.0 corrected) were only at chance. With task-based fMRI data, the automated classifier performed at chance in all three severity groups. CONCLUSIONS: Binary linear SVM classifiers achieved significant classification of very severe depression with resting-state fMRI, but the contribution of brain measurements may have limited potential in differentiating patients with less severe depression from healthy controls.

A preliminary study of the influence of age of onset and childhood trauma on cortical thickness in major depressive disorder.

BACKGROUND: Major depressive disorder (MDD) neural underpinnings may differ based on onset age and childhood trauma. We assessed cortical thickness in patients who differed in age of MDD onset and examined trauma history influence. METHODS: Adults with MDD (N=36) and controls (HC; N=18) underwent magnetic resonance imaging. Twenty patients had MDD onset<24 years of age (pediatric onset) and 16 had onset>25 years of age (adult onset). The MDD group was also subdivided into those with (N=12) and without (N=19) physical and/or sexual abuse as assessed by the Childhood Trauma Questionnaire (CTQ). Cortical thickness was analyzed with FreeSurfer software. RESULTS: Thicker frontal pole and a tendency for thinner transverse temporal cortices existed in MDD. The former was driven by the pediatric onset group and abuse history (independently), particularly in the right frontal pole. Inverse correlations existed between CTQ scores and frontal pole cortex thickness. A similar inverse relation existed with left inferior and right superior parietal cortex thickness. The superior temporal cortex tended to be thinner in pediatric versus adult onset groups with childhood abuse. CONCLUSIONS: This preliminary work suggests neural differences between pediatric and adult MDD onset. Trauma history also contributes to cytoarchitectural modulation. Thickened frontal pole cortices as a compensatory mechanism in MDD warrant evaluation.

Influence of age of onset on limbic and paralimbic structures in depression.

AIM: Major depressive disorder (MDD) onset during childhood/adolescence is associated with a greater illness burden and distinct clinical profile. However, limited research exists on the effect of age of MDD onset on volumetric abnormalities in para/limbic structures during adulthood. METHODS: Subgenual anterior cingulate cortex (sgACC), hippocampus and caudate nucleus volumes were measured by manual tracing in depressed individuals (n = 45) and healthy controls (HC; n = 19). Volumetric comparisons were carried out between HC and MDD patients divided into those with pediatric (≤ 18 years; n = 17) and adult onset (≥ 19 years; n = 28). RESULTS: The adult MDD-onset group had smaller sgACC volumes than the pediatric-onset and HC groups (age, sex controlled). No differences in caudate and hippocampus volumes existed. sgACC and hippocampal volumes were inversely correlated with depression severity. CONCLUSIONS: Surprisingly, pediatric MDD-onset was not associated with more pronounced sgACC, hippocampus and caudate volume reductions. Nevertheless, age of illness onset appears to be a meaningful dimension of study in efforts to understand the neurobiological heterogeneity of MDD.

Differential neural activity and connectivity for processing one's own face: a preliminary report.

The experience of self is unique and pivotal to clinically relevant cognitive and emotional functions. However, well-controlled data on specialized brain regions and functional networks underlying the experience of self remain limited. This functional magnetic resonance imaging study investigated neural activity and connectivity specific to processing one's own face in healthy women by examining neural responses to the pictures of the subjects' own faces in contrast to faces of their own mothers, female friends and strangers during passive viewing, emotional and self-relevance evaluations. The processing of one's own face in comparison to processing of familiar faces revealed significant activity in right anterior insula (AI) and left inferior parietal lobule (IPL), and less activity in right posterior cingulate/precuneus (PCC/PCu) across all tasks. Further, the seed-based correlation analysis of right AI, and left IPL, showed differential functional networks in self and familiar faces contrasts. There were no differences in valence and saliency ratings between self and familiar others. Our preliminary results suggest that the self-experience cued by self-face is processed predominantly by brain regions and related networks that link interoceptive feelings and sense of body ownership to self-awareness and less by regions of higher order functioning such as autobiographical memories.