Search

The promise of machine learning has fueled the hope for developing diagnostic tools for psychiatry. Initial studies showed high accuracy for the identification of major depressive disorder (MDD) with resting-state connectivity, but progress has been hampered by the absence of large datasets. Here we used regular machine learning and advanced deep learning algorithms to differentiate patients with MDD from healthy controls and identify neurophysiological signatures of depression in two of the largest resting-state datasets for MDD. We obtained resting-state functional magnetic resonance imaging data from the REST-meta-MDD (N = 2338) and PsyMRI (N = 1039) consortia. Classification of functional connectivity matrices was done using support vector machines (SVM) and graph convolutional neural networks (GCN), and performance was evaluated using 5-fold cross-validation. Features were visualized using GCN-Explainer, an ablation study and univariate t-testing. The results showed a mean classification accuracy of 61% for MDD versus controls. Mean accuracy for classifying (non-)medicated subgroups was 62%. Sex classification accuracy was substantially better across datasets (73-81%). Visualization of the results showed that classifications were driven by stronger thalamic connections in both datasets, while nearly all other connections were weaker with small univariate effect sizes. These results suggest that whole brain resting-state connectivity is a reliable though poor biomarker for MDD, presumably due to disease heterogeneity as further supported by the higher accuracy for sex classification using the same methods. Deep learning revealed thalamic hyperconnectivity as a prominent neurophysiological signature of depression in both multicenter studies, which may guide the development of biomarkers in future studies.

Objective: Electroconvulsive therapy (ECT) is the most effective treatment for patients with severe major depressive disorder (MDD). Given the known sex differences in MDD, improved knowledge may provide more sex-specific recommendations in clinical guidelines and improve outcome. In the present study we examine sex differences in ECT outcome and its predictors. Methods: Clinical data from 20 independent sites participating in the Global ECT-MRI Research Collaboration (GEMRIC) were obtained for analysis, totaling 500 patients with MDD (58.6 % women) with a mean age of 54.8 years. Severity of depression before and after ECT was assessed with validated depression scales. Remission was defined as a HAM-D score of 7 points or below after ECT. Variables associated with remission were selected based on literature (i.e. depression severity at baseline, age, duration of index episode, and presence of psychotic symptoms). Results: Remission rates of ECT were independent of sex, 48.0 % in women and 45.7 % in men (X2(1) = 0.2, p = 0.70). In the logistic regression analyses, a shorter index duration was identified as a sex-specific predictor for ECT outcome in women (X2(1) = 7.05, p = 0.01). The corresponding predictive margins did show overlapping confidence intervals for men and women. Conclusion: The evidence provided by our study suggests that ECT as a biological treatment for MDD is equally effective in women and men. A shorter duration of index episode was an additional sex- specific predictor for remission in women. Future research should establish whether the confidence intervals for the corresponding predictive margins are overlapping, as we find, or not.

Major depressive disorder (MDD) is associated with abnormal neural circuitry. It can be measured by assessing functional connectivity (FC) at resting-state functional MRI, that may help identifying neural markers of MDD and provide further efficient diagnosis and monitor treatment outcomes. The main aim of the present study is to investigate, in an unbiased way, functional alterations in patients with MDD using a large multi-center dataset from the PsyMRI consortium including 1546 participants from 19 centers ( www.psymri.com ). After applying strict exclusion criteria, the final sample consisted of 606 MDD patients (age: 35.8 ± 11.9 y.o.; females: 60.7%) and 476 healthy participants (age: 33.3 ± 11.0 y.o.; females: 56.7%). We found significant relative hypoconnectivity within somatosensory motor (SMN), salience (SN) networks and between SMN, SN, dorsal attention (DAN), and visual (VN) networks in MDD patients. No significant differences were detected within the default mode (DMN) and frontoparietal networks (FPN). In addition, alterations in network organization were observed in terms of significantly lower network segregation of SMN in MDD patients. Although medicated patients showed significantly lower FC within DMN, FPN, and SN than unmedicated patients, there were no differences between medicated and unmedicated groups in terms of network organization in SMN. We conclude that the network organization of cortical networks, involved in processing of sensory information, might be a more stable neuroimaging marker for MDD than previously assumed alterations in higher-order neural networks like DMN and FPN.

Background: Trauma-focused psychotherapy is a first-line treatment for PTSD but 30–50% of patients do not benefit sufficiently (Bradley et al., 2005). Neuroimaging has been proposed as a potential biomarker predicting treatment-response in PTSD patients (Colvonen et al., 2017; Yuan et al., 2018). Objective: We investigated whether neuroimaging data could distinguish between treatment responders and non-responders on the group and single-subject level. Method: A total of 44 male veterans with PTSD underwent baseline structural and resting-state MRI followed by trauma-focused therapy (EMDR or TFCBT). Grey-matter volumes (GMV) were extracted from the MRI data and resting-state networks (RSN) were estimated using group-ICA of data from 28 matched trauma-exposed healthy controls. GMV and RSNs were used to find differences between responders and non-responders on the group and single-subject level. Treatment response was defined as 30% decrease in total Clinician-Administered PTSD Scale for the DSM-IV (CAPS-IV) score from pre- to post-treatment assessment. Gaussian process classifiers with 10 times repeated 10-fold cross-validation were used for classification. Results: An RSN centred on the pre-SMA could distinguish between responders and non-responders on an individual level with 81.4% accuracy (p < .001), 84.5% sensitivity, 78% specificity and AUC of 0.93, while an RSN centred on the bilateral superior frontal gyrus differed between groups (pFWE < .05). No significant singlesubject classification or group differences were 54 ESTSS 2019 Rotterdam Symposium Abstract Book observed for GMV. Conclusions: Rs-fMRI is capable of providing personalized predictions of treatment response in a sample of veterans with PTSD. It therefore has the potential to be useful as a biomarker of treatment response.

BACKGROUND: The ventral anterior limb of the internal capsule (vALIC) is a target for deep brain stimulation (DBS) in obsessive-compulsive disorder (OCD). Conventional surgical planning is based on anatomical landmarks.OBJECTIVE/HYPOTHESIS: We hypothesized that treatment response depends on the location of the active DBS contacts with respect to individual white matter bundle trajectories. This study thus aimed to elucidate whether vALIC DBS can benefit from bundle-specific targeting.METHODS: We performed tractography analysis of two fiber bundles, the anterior thalamic radiation (ATR) and the supero-lateral branch of the medial forebrain bundle (MFB), using diffusion-weighted magnetic resonance imaging (DWI) data. Twelve patients (10 females) who had received bilateral vALIC DBS for at least 12 months were included. We related the change in OCD symptom severity on the Yale-Brown obsessive-compulsive scale (Y-BOCS) between baseline and one-year follow-up with the distances from the active contacts to the ATR and MFB. We further analyzed the relation between treatment response and stimulation sites in standard anatomical space.RESULTS: We found that active stimulation of the vALIC closer to the MFB than the ATR was associated with better treatment outcome (p = 0.04; r2 = 0.34). In standard space, stimulation sites were largely overlapping between treatment (non)responders, suggesting response is independent of the anatomically defined electrode position.CONCLUSION: These findings suggest that vALIC DBS for OCD may benefit from MFB-specific implantation and highlight the importance of corticolimbic connections in OCD response to DBS. Prospective investigation is necessary to validate the clinical use of MFB targeting.

"Slips of action" occur in everyday life when we momentarily lose sight of a goal (for example, when in a rush or distracted). Associative models propose that these habitual responses can be activated via a direct stimulus-response (S-R) mechanism, regardless of the current hedonic value of the outcome. The slips-of-action task (SOAT) has been extensively used in both healthy and pathological populations to measure habit tendencies, the likelihood of making erroneous responses for devalued outcomes. Inspection of behavioral performance does not reveal, however, whether the impairments were due to impaired goal-directed control or aberrantly strong habit formation. In the current study, we used functional MRI while human participants performed both the instrumental training and SOAT test phases, to elucidate the relative contributions of these mechanisms to performance on the SOAT. On trials in which conflict arises between competing goal-directed and habitual responses, we observed increased activation across areas including the anterior cingulate cortex, paracingulate gyrus, lateral orbitofrontal cortex (OFC), insula, and inferior frontal gyrus (IFG). Responding for devalued outcomes was related to increased activation in the premotor cortex and cerebellum, implicating these regions in habitual responding. Increased activation in the caudate, dorsolateral prefrontal cortex (dlPFC), and frontal pole during training was associated with better performance during the test phase, indicative of goal-directed action control. These results endorse interpretation of the SOAT in terms of competing goal-directed and habitual mechanisms and highlight that cognitive control processes present an additional bottleneck for successful performance on this task.

Glutamate and GABA play an important role in substance dependence. However, it remains unclear whether this holds true for different substance use disorders and how this is related to risk-related traits such as impulsivity. We, therefore, compared Glx (as a proxy measure for glutamate) and GABA concentrations in the dorsal anterior cingulate cortex (dACC) of 48 male cigarette smokers, 61 male smoking polysubstance users, and 90 male healthy controls, and investigated the relationship with self-reported impulsivity and substance use. Glx and GABA concentrations were measured using proton Magnetic Resonance Spectroscopy. Impulsivity, smoking, alcohol and cocaine use severity and cannabis use were measured using self-report instruments. Results indicate a trend towards group differences in Glx. Post-hoc analyses showed a difference between smokers and healthy controls (p=0.04) and a trend towards higher concentrations in smoking polysubstance users and healthy controls (p=0.09), but no differences between smokers and smoking polysubstance users. dACC GABA concentrations were not significantly different between groups. Smoking polysubstance users were more impulsive than smokers, and both groups were more impulsive than controls. No significant associations were observed between dACC neurotransmitter concentrations and impulsivity and level and severity of smoking, alcohol or cocaine use or the presence of cannabis use. The results indicate that differences in dACC Glx are unrelated to type and level of substance use. No final conclusion can be drawn on the lack of GABA differences due to assessment difficulties. The relationship between dACC neurotransmitter concentrations and cognitive impairments other than self-reported impulsivity should be further investigated.

OBJECTIVE: The authors investigated whether cocaine use disorder is associated with abnormalities in the neural underpinnings of aversive conditioning and extinction learning, as these processes may play an important role in the development and persistence of drug abuse.METHOD: Forty male regular cocaine users and 51 male control subjects underwent a fear conditioning and extinction protocol during functional MRI. Skin conductance response was measured throughout the experiment as an index of conditioned responses.RESULTS: Cocaine users showed hyperresponsiveness of the amygdala and insula during fear conditioning, as well as hyporesponsiveness of the dorsomedial prefrontal cortex during extinction learning. In cocaine users, but not in control subjects, skin conductance responses were positively correlated with responsiveness of the insula, amygdala, and dorsomedial prefrontal cortex during fear conditioning but negatively correlated with responsiveness of the ventromedial prefrontal cortex during extinction learning.CONCLUSIONS: Increased sensitivity to aversive conditioned cues in cocaine users might be a risk factor for stress-relief craving in cocaine use disorder. These results support the postulated role of altered aversive conditioning in cocaine use disorder and may be an important step in understanding the role of aversive learning in the pathology of cocaine use disorder.

Appetitive conditioning refers to the process of learning cue-reward associations and is mediated by the mesocorticolimbic system. Appetitive conditioned responses are difficult to extinguish, especially for highly salient reward such as food and drugs. We investigate whether aversive counterconditioning can alter reward reinstatement in the ventral striatum in healthy volunteers using functional magnetic resonance imaging (fMRI). In the initial conditioning phase, two different stimuli were reinforced with a monetary reward. In the subsequent counterconditioning phase, one of these stimuli was paired with an aversive shock to the wrist. In the following extinction phase, none of the stimuli were reinforced. In the final reinstatement phase, reward was reinstated by informing the participants that the monetary gain could be doubled. Our fMRI data revealed that reward signaling in the ventral striatum and ventral tegmental area following reinstatement was smaller for the stimulus that was counterconditioned with an electrical shock, compared to the non-counterconditioned stimulus. A functional connectivity analysis showed that aversive counterconditioning strengthened striatal connectivity with the hippocampus and insula. These results suggest that reward signaling in the ventral striatum can be attenuated through aversive counterconditioning, possibly by concurrent retrieval of the aversive association through enhanced connectivity with hippocampus and insula.