Toward literature-based feature selection for diagnostic classification: a meta-analysis of resting-state fMRI in depression

Abstract

Information derived from functional magnetic resonance imaging (fMRI) during wakeful rest has been introduced as a candidate diagnostic biomarker in unipolar major depressive disorder (MDD). Multiple reports of resting state fMRI in MDD describe group effects. Such prior knowledge can be adopted to pre-select potentially discriminating features for diagnostic classification models with the aim to improve diagnostic accuracy. Purpose of this analysis was to consolidate spatial information about alterations of spontaneous brain activity in MDD, primarily to serve as feature selection for multivariate pattern analysis techniques (MVPA). Thirty two studies were included in final analyses. Coordinates extracted from the original reports were assigned to two categories based on directionality of findings. Meta-analyses were calculated using the non-additive activation likelihood estimation approach with coordinates organized by subject group to account for non-independent samples. Converging evidence revealed a distributed pattern of brain regions with increased or decreased spontaneous activity in MDD. The most distinct finding was hyperactivity/hyperconnectivity presumably reflecting the interaction of cortical midline structures (posterior default mode network components including the precuneus and neighboring posterior cingulate cortices associated with self-referential processing and the subgenual anterior cingulate and neighboring medial frontal cortices) with lateral prefrontal areas related to externally-directed cognition. Other areas of hyperactivity/hyperconnectivity include the left lateral parietal cortex, right hippocampus and right cerebellum whereas hypoactivity/hypoconnectivity was observed mainly in the left temporal cortex, the insula, precuneus, superior frontal gyrus, lentiform nucleus and thalamus. Results are made available in two different data formats to be used as spatial hypotheses in future studies, particularly for diagnostic classification by MVPA.

Time related effects on functional brain connectivity after serotonergic and cholinergic neuromodulation

“…This does not imply that the partial (time specific) effects are meaningless. A decrease in connectivity at T  = 4.5 h between the default mode network and precuneus, PCC and ACC is in line with earlier results [Klaassens et al, 2015; McCabe and Mishor, 2011; McCabe et al, 2011; Van de Ven et al, 2013; Van Wingen et al, 2014] and in agreement with opposite features in depression, which is characterized by increased connectivity of DMN components [Sundermann et al, 2014]. Especially the posterior part of the DMN, where citalopram effects were most prevalent, has been implicated in SSRI efficacy in depression [Greicius et al, 2007; Li et al, 2013].…”

“…Compared with our recent results on the SSRI sertraline [Klaassens et al, 2015], there was considerable overlap between the two SSRIs in direction (decreased connectivity) and regions (ACC, PCC, precuneus, prefrontal cortex, midbrain, and motor cortex) of effect, especially with respect to other pharmacological compounds that usually show more restricted responses [Khalili‐Mahani et al, 2012; Klumpers et al, 2012; Niesters et al, 2012]. Part of these findings is in line with RS‐fMRI studies in depressed patients who exhibit hyperconnectivity of cortical midline structures (ACC, PCC, precuneus, and medial prefrontal regions) that are related to emotion regulation and modulated by serotonin transmission [Kupfer et al, 2012; Sundermann et al, 2014]. It has been hypothesized that this increase in connectivity in depression is representative of disruptions in self‐consciousness and rumination of negative thoughts [Hamilton et al, 2011; Zhu et al, 2012].…”

“…Especially the posterior part of the DMN, where citalopram effects were most prevalent, has been implicated in SSRI efficacy in depression [Greicius et al, 2007; Li et al, 2013]. Furthermore, consistent with an increased cerebellar‐DMN connectivity in depression [Sundermann et al, 2014], citalopram reduced connectivity between the DMN and cerebellum. The cerebellum is primarily known for its service in motor control, illustrating our findings for the sensorimotor network, but influences mood regulation as well [Schmahmann, 2004; Schmahmann and Sherman, 1998].…”

“…These networks have consistently been found in healthy and clinical conditions, and are related to specific functions of the brain (i.e., motor, auditory, visual, emotional, and executive function) [Beckmann et al, 2005; Damoiseaux et al, 2006; Seeley et al, 2007; Smith et al, 2009]. Disruptions of functional networks have been demonstrated in both depressed and demented patients, especially for the default mode network (DMN) [Hafkemeijer et al, 2012; Sundermann et al, 2014; Wang et al, 2006]. Several studies point to normalization of DMN connectivity in depression after SSRI administration [McCabe and Mishor, 2011; McCabe et al, 2011; Van de Ven et al, 2013; Van Wingen et al, 2014].…”

Orbital and Medial Prefrontal Cortex Functional Connectivity of Major Depression Vulnerability and Disease

“…On the theoretical side, neuro-circuit models of MD propose altered pathways based only on tracer or lesion findings in animals and rare post mortem human reports (13). Thus far, MRI studies of MD have examined functional wiring of the OMPFC either only for select coordinates derived from activation studies or coarsely, as either part of the anterior default mode network (DMN) or with predefined general anatomical labels (7–10; see reviews in 11,12). …”