Skip to main content



Project 1: The neural network OCD, linking NHB and human studies

The goal of P1 is to gain insight into the anatomical underpinnings for connectivity abnormalities seen in Obsessive Compulsive Disorder (OCD). The pathophysiology of OCD is associated with dysfunction in prefrontal cortico-basal ganglia circuits, in particular the dorsal anterior cingulate cortex (dACC), orbital frontal cortex (OFC), and dorsal striatum (DS). These areas are linked to regions involved in both emotion and higher cognitive and motor control. We hypothesize that within the dACC, OFC, vlPFC, and striatum, there are specific regions (referred to as hubs) that are interconnected and receive convergent inputs from each other, from the amygdala and/or ventromedial prefrontal cortex (vmPFC), and from the dorsolateral PFC (dlPFC), and/or pre supplementary motor area (SMA). These hubs would provide an anatomical substrate for modulation between emotional and cognitive systems. The cingulum bundle, corpus callosum, and uncinate fasciculus, are the white matter (WM) bundles that interconnect these cortical areas, and, along with the internal capsule, also show abnormalities in OCD compared to healthy controls. We also hypothesize that abnormalities in these bundles will be prominent in the WM connecting the critical nodes. Finally, we predict that neurostimulation and cingulotomy will affect WM fractional anisotropy, radial diffusivity (FA/RD) and volume. Our three aims are: 1. To identify and characterize hubs in regions of the dACC, vlPFC, OFC, and striatum; 2. Characterize the organization of WM that connects the OCD circuit; and 3. Determine if differences in connectivity in OCD patients are associated with WM segments.

Project 2: Characterizing abnormalities in the OCD neural network

The goal of Project 2 (P2) is to identify intrinsic (resting state) functional and white matter (WM) abnormalities in a neural network consistently implicated in OCD. This network includes two key cortical regions, the orbital frontal cortex (OFC) and the dorsal anterior cingulate cortex (dACC), that link motivation with the ability to adapt behaviors based on perceived outcomes. Our overarching hypothesis is that in OCD there is abnormally increased positive connectivity between OFC/dACC and emotion processing areas, that include the amygdala (Amyg) and ventral medial prefrontal cortex (vmPFC), and decreased connectivity between OFC/dACC and the ventral lateral prefrontal cortex (vlPFC) and dorsolateral prefrontal cortex (dlPFC), implicated in cognitive control. This change in balance between emotional processing and cognitive control systems results in OFC/dACC excessively driving the dorsal striatum (DS), which may trigger the excessive ritualized behaviors in seen OCD. Specifically, we plan to 1) Characterize in OCD relative to healthy individuals resting state connectivity among dACC, OFC, vlPFC, dlPFC, vmPFC, Amyg, and DS; 2) Characterize in OCD relative to healthy individuals with dMRI and probabilistic tractography the integrity of key WM bundles; and 3) Integrate neuroimaging measures to identify relationships between abnormalities in resting state connectivity (A1) and WM (A2) in individuals with OCD and healthy individuals.

Project 3: Targeting neurocircuitry relevant to OCD in individuals

The goal of P3 is to explore methods for identifying OCD-relevant circuitry in individuals. As targets for invasive and non-invasive OCD treatments become available, a bottleneck for translation in clinical practice is the ability to robustly identify affected circuits in individual patients and measure network modulation. More broadly, focus on groups in neuroimaging studies has slowed translation of findings to the clinical arena where the focus is on the individual. If successful, the project will provide a general set of methods for identifying within-individual network measures that can guide interventions and serve as tailored biomarkers for clinical readouts. The central hypothesis of our Center is that OCD is characterized by hyperconnectivity between the amyg/vmPFC and dACC/ OFC and decreased connectivity between dlPFC/vlPFC and dACC/OFC. This change in balance between emotional processing and cognitive control systems results in a pathological heightened activity state in the dACC/OFC/vlPFC/DS circuitry. In this project we will explore core nodes of this circuitry within individuals and use these methods to explore neuromodulatory effects of OCD circuitry. Aim 1. Identify OCD-relevant circuitry in healthy individuals from the available data including the spatial variability in node location. We hypothesize that, while clustering around the group central tendency, the location of key nodes including the dACC and vlPFC will spatially shift across individuals. These measures will be used to build a variability map. Aim 2. Develop and validate an HCP-inspired connectomic acquisition and analysis strategy targeting diffusion and functional connectivity on a widely available MRI platform. Aim 3. Examine changes in OCD circuitry before and after neuromodulation, 3A. OCD circuitry will be measured in 12 individuals before and after cingulotomy. We hypothesize that coupling between DS and regions just rostral to the cingulotomy will be reduced in effective treatment and preserved in ineffective treatment. 3B. A within-subject TMS study motivated by P4 will analyze before vs after TMS stimulation to further explore whether connectivity measures can be used as a readout of circuit-level perturbations in the individual.

Project 4: Modulation of dACC/pSMA by tDCS, and by nonfocal and focal rTMS in OCD

The Obsessive-Compulsive Disorder (OCD) Clinic and Research Program at Butler Hospital and the Alpert Medical School of Brown University has been a leading center in research and treatment of OCD for three decades. A major focus of research is developing treatments based on changing the functioning of brain circuits involved in the symptoms of OCD and related conditions. The OCD Clinic leads a national study of deep brain stimulation for highly selected individuals with severely disabling (“intractable”) OCD, and continues to provide neurosurgical ablation as an alternative for this group of patients. Most recently, the Clinic has focused on research using noninvasive methods of brain stimulation, which can be applied to a larger group of individuals suffering from OCD. These methods deliver electrical or magnetic stimulation through the scalp and skull: transcranial direct current stimulation and transcranial magnetic stimulation. Current projects combine one of these stimulation techniques with behavior therapy, or, as s part of this NIMH Conte Center, with neuroimaging.

Project 5: Cingulate / prelimbic nodes mediating persistent avoidance

In Project 5, we are using a rodent model of OCD-like behavior, persistent active avoidance, in which rats continue to avoid a tone for which the conditioned association has been extinguished. Using optogenetic tools, we are investigating the role of the prelimbic-cingulate cortex (PL/Cg) in this behavior, as well as inputs to PL/Cg from orbitofrontal cortex and amygdala. Our hypothesis is that connection hubs within PL/Cg mediate persistent avoidance, and are homologous to human cortical areas implicated in OCD.

Core B: Anatomy

The overall goal of this Center is to understand the neural network associated with obsessive-compulsive disorder (OCD). The data collected by Center investigators comes from a wide range of sources including studies of rodents, primates and humans. One of the major aims of the Center will be to broaden our understanding of the functional neurocircuitry of the anterior cingulate, orbital, and ventrolateral prefrontal cortices as they relate to abnormalities in OCD. The overall mission of the anatomy core is to provide the bridge between projects to identify structures that are likely to be affected by the disorder and modulated by neural stimulation across species. To achieve this goal, the anatomy cores work closely with each of the individual projects and Core C, to develop cross-species models of structures and pathways. The first aim of Core B is to provide support for the animal anatomic by assisting in tissue-processing in developing models of the cortical and striatal hubs across species. The second aim of core B is to provide support for dMRI studies that require assistance in acquisition, processing, and data analysis. Core B is responsible for collecting and processing diffusion dMRI data and providing analysis support and data sharing for projects.

Core C: Neuroimaging acquisition and neuroinformatics

The experiments proposed in the Center include extensive use of neuroimaging in both normal control participants and patient groups. Core C will support the individual projects in data acquisition and quality control, implementation of processing pipelines, and capture of raw data for cross-project analyses and eventual broader distribution to the community. Aim 1: The core will assist in implementation of MRI acquisition procedures and anticipated transitions in MR hardware. All project sites currently have available similar Siemens Trio scanners and anticipate transition to the PRISMA system in coming years. The uniformity of hardware and expected transitions provides an opportunity to unify data acquisition and facilitate transition to next-generation functional MRI data acquisition procedures. Optimization of target PRISMA sequences will be performed including direct analysis of comparability to existing Trio sequences. Aim 2: A centralized database will capture all neuroimaging data collected across sites. The neuroinformatics backbone will be based on a custom implementation of the Extensible Neuroimaging Archive Toolkit (XNAT; Marcus et al., 2007). The installation will store each subject’s data with metadata about the data types, etc. to allow future uses. These data will be accessible to project investigators and stored ready for wide data sharing. Aim 3: The core will provide quality control data assessments. Data quality and uniformity is important to MRI data acquisition, in particular for functional MRI data that is notorious plagued by artifacts of subject compliance and head motion. The core will assess all neuroimaging data acquired in P1-4 for artifacts and make recommendations to improve data quality. Quantitative quality control measures will include assessment of motion and overall signal-to-noise characteristics. The quality control procedures are modeled after those adopted by the NIMH EMBARC trial and implemented in the Simons VIP projects studying autism. Aim 4: The core will develop novel statistical models and computational tools for an fcMRI test for differences between populations (OCD vs healthy controls), effects of treatment (cingulotomy; TMS; direct and indirect modulation) while accounting for between-subject variation.