Studying the Relapsing Dynamics of Major Depression through Network Analysis of Fmri Connectivity Maps: Implications for Therapy

Subproject 1: Main objective: The objective of this study was to analyse the network of the functional connectivity map supporting mood disorders: major depressive disorder (MDD) and bipolar disease (BD). The specific aims were: 1) to reverse engineer the functional connectivity map supporting MDD and BD; 2) to develop computational models of the functional connectivity maps of MDD for identifying the control elements driving the relapsing dynamics in mood disorders; 3) to validate the networks models of mood disorders through computational model simulations and perturbations. Methodology: i) Methods: 1. Patients: cohort of patients with MDD and BD according to DMS-IV criteria and matched controls. Patients were collected at the time of suffering a new episode of depression; 2. fMRI studies: resting state, mood-related tasks (sad faces and sadness evocation) and working memory task, diffusion spectra imaging (DSI). Analysis of functional connectivity maps was done using the SPM8 connectivity toolbox and analysis of structural connectivity maps (tractography) was done using the connectome mapper. 4. Computational modelling: network models for topological analysis and ordinary differential equation models for dynamic analysis; 5. Analysis of the network dynamics using dynamic theory. ii) Design: Aim 1: In order to obtain the mood functional connectivity map, fMRI from resting state, mood-related task and working memory tasks were obtained in patients with MDD or BD. Available information in the literature and fMRI databases was used to provide a template of the network. Aim 2: The mood disorder network modelling was focused on the identification of control mechanisms that allow the generation of relapses (i.e. negative feedback). We generated mathematical models and performed sensitivity analysis in order to identify the critical components involved. Aim 3: Predictions from computational models were validated in computational simulations of additional patient datasets by assessing changes in the MDD network. 3 Subproject 2: As stated in the original project proposal, recent functional magnetic resonance imaging (fMRI) studies identified components of the neural networks that participate in the generation of depressive mood. The objective of this study was to analyse under resting state conditions, i.e. in the absence of a task, the network and control properties of the functional connectivity map supporting depressive symptoms. This approach showed promise in leading to the discovery of novel useful diagnostic markers. Our specific goal was to model the subjects' brain connectivity in order to perform topological as well as dynamic analyses. Using our models we approximated resting …