Date created: | Last Updated:

Category: Project

Description: This study will assess differences in resting state functional magnetic resonance imaging (fMRI) Dynamic Causal Modeling-estimated effective connectivity between cocaine dependent (CD) subjects and healthy control (HC) subjects. Effective connectivity between-brain networks estimated by independent component analysis (ICA) will be conducted. Networks to be examined include the default mode network (DMN), salience network (SN), left executive control network (LECN), and right executive control network (RECN). The subject sample will include resting state fMRI data from 22 CD subjects with DSM-IV-diagnosed CD and 22 HC subjects compiled from three studies. This sample of 22 CD and 22 HC subjects was refined from a sample of 28 CD and 28 HC subjects by more closely matching the two groups for age and years of education attained. We will also compare the unrefined 28 CD and 28 HC sample in a supplementary analysis. All subjects in both groups met Parkes et al. (2018) stringent criteria for head motion and were right-handed. A preliminary functional connectivity analysis was conducted as described in Woisard et al. (2022). The output of that functional connectivity analysis will serve as the input for the present Dynamic Causal Modeling analysis. In order to address the effects of recent cocaine and cannabis use within the CD group, we will also include whether the CD subject’s urine drug screen the day of their MRI scan was positive for cocaine or cannabis as factors in a general linear model. We will also compare females and males in the analysis. The fMRI acquisition parameters were: repetition time 1500 ms, echo time 30 ms, flip angle 68°, field of view 240 mm (anterior-to-posterior) × 240 mm (left-to-right) × 143.67 mm (foot-to-head), in-plane resolution 3.75 mm × 3.75 mm, 32 axial slices, slice thickness 3.75 mm, interslice gap 0.76 mm, 375 repetitions per run after 12 dummy acquisitions, and total duration 9 minutes. Similar to Woisard et al. (2021), initial removal of signal outliers, slice timing correction, spatial smoothing, and registration to a T1-weighted anatomical scan were performed. Quality control for head motion was performed by eliminating fMRI runs which did not meet the Parkes et al. (2018) stringent criteria. Head motion re-alignment and signal correction was performed using the FSL MCFLIRT motion-correction program (Jenkinson et al., 2002; Woolrich et al., 2004) and ICA-AROMA (Pruim et al., 2015), respectively. Further denoising was performed using the aCompCor procedure implemented in CONN software (Behzadi et al., 2007; www.nitrc.org/projects/conn, RRID:SCR_009550), and ICA components with possible motion-related structured noise are regressed out during the FSL dual regression procedure (Bijsterbosch et al., 2017, pp. 64-65). The T1-weighted anatomical scan was transformed into MNI space using the FSL non-linear transformation module FNIRT, after which these parameters were applied to the denoised fMRI timeseries for transformation into MNI space. High pass filtering (cutoff=125 s), but not low pass filtering, was performed. ICA was conducted with FSL Multivariate Exploratory Linear Optimized Decomposition into Independent Components (MELODIC; Beckmann & Smith, 2004; https://fsl.fmrib.ox.ac.uk/fsl/fslwiki/MELODIC) with a dimensionality of 30, in order to match our previously conducted study in opioid use disorder (Woisard et al., 2021). The output was inspected visually to identify the DMN, SN, LECN, and RECN, based on previous published studies (Menon, 2011; Shirer et al., 2012; Sridharan et al., 2008; Woisard et al., 2021). The dual regression procedure (Nickerson et al., 2017) implemented in FSL was then used to generate a subject-specific timecourse for each of the DMN, SN, LECN, and RECN network components. These subject-specific timeseries Dual Regression stage 1 outputs for the SN, DMN, left ECN, and right ECN will be used as the input for the Dynamic Causal Modeling effective connectivity analysis. Stochastic Dynamic Causal Modeling will be conducted to assess between-network effective connectivity as well as for the presence of non-linear modulatory effects of a given network on the effective connectivities between other networks. We will first assess whether the Goulden et al. (2014) findings of the SN modulating the effective connectivity between the DMN and ECN can be replicated in HC and whether these findings differ between HC and CD. We will accomplish this by estimating and comparing four models for each subject, each with intrinsic connections between each of the SN, DMN, left ECN, and right ECN, and with no extrinsic inputs to the model. The models will test which network modulates the connections between the other networks (i.e., one model will entail modulation by the SN of the connections between the DMN and left ECN and right ECN, and so on for each network in turn as modulator of the others). The “best” model will be determined by Bayesian model selection (Stephan et al., 2009), following Goulden et al. (2014). We will then compare the effective connectivities across subjects using the Parametric Empirical Bayes approach. Relative to the classic methods (Stephan et al., 2009; Penny et al., 2010) for evaluating group difference in Dynamic Causal Modeling, the results from the Parametric Empirical Bayesian approach could be more accurate (Friston et al., 2016) because all the single subject dynamic causal models can be iteratively re-estimated using updated Bayesian priors of the averaged connectivity parameters at the group level. The Bayesian Posterior Probabilities will be reported for the group differences in connectivity, which will be calculated by the Parametric Empirical Bayes procedure. We hypothesize the dynamic causal model of the SN modulating the effective connectivity between the DMN and ECN will be the “best” model among those tested in HC, as found in the Goulden et al. (2014) study in healthy control subjects, and this finding will be replicated in CD. We also hypothesize there will be differences in effective connectivity between the ECN and DMN, ECN and SN, and SN and DMN in CD compared to HC. We will also examine self-connectivity differences between groups for each network in exploratory fashion (Snyder et al., 2021).   References Beckmann, C. F., & Smith, S. M. (2004). Probabilistic independent component analysis for functional magnetic resonance imaging. IEEE transactions on medical imaging, 23(2), 137-152. Behzadi, Y., Restom, K., Liau, J., & Liu, T. T. (2007). A component based noise correction method (CompCor) for BOLD and perfusion based fMRI. Neuroimage, 37(1), 90-101. Bijsterbosch, J., Smith, S. M., & Beckmann, C. F. (2017). An introduction to resting state fMRI functional connectivity (pp. 59, 63-64, 67, 75). Oxford University Press. Friston, K. J., Litvak, V., Oswal, A., Razi, A., Stephan, K. E., Van Wijk, B. C., ... & Zeidman, P. (2016). Bayesian model reduction and empirical Bayes for group (DCM) studies. Neuroimage, 128, 413-431. Goulden, N., Khusnulina, A., Davis, N. J., Bracewell, R. M., Bokde, A. L., McNulty, J. P., & Mullins, P. G. (2014). The salience network is responsible for switching between the default mode network and the central executive network: replication from DCM. Neuroimage, 99, 180-190. Jenkinson, M., Bannister, P., Brady, M., & Smith, S. (2002). Improved optimization for the robust and accurate linear registration and motion correction of brain images. Neuroimage, 17(2), 825-841. Kutner, M. H., Nachtsheim, C. J., Neter, J., & Li, W. (2005). Applied linear statistical models (Vol. 5; pp. 347, 919, 940). Boston: McGraw-Hill Irwin. Ma, L., Hettema, J. M., Cousijn, J., Bjork, J. M., Steinberg, J. L., Keyser-Marcus, L., Woisard, K., Lu, Q., Roberson-Nay, R., Abbate, A., & Moeller, F. G. (2020). Resting-state directional connectivity and anxiety and depression symptoms in adult cannabis users. Biological Psychiatry: Cognitive Neuroscience and Neuroimaging. Menon, V. (2011). Large-scale brain networks and psychopathology: a unifying triple network model. Trends in cognitive sciences, 15(10), 483-506. Nickerson, L. D., Smith, S. M., Öngür, D., & Beckmann, C. F. (2017). Using dual regression to investigate network shape and amplitude in functional connectivity analyses. Frontiers in neuroscience, 11, 115. Penny, W. D., Stephan, K. E., Daunizeau, J., Rosa, M. J., Friston, K. J., Schofield, T. M., & Leff, A. P. (2010). Comparing families of dynamic causal models. PLoS Comput Biol, 6(3), e1000709. Pruim, R. H., Mennes, M., Buitelaar, J. K., & Beckmann, C. F. (2015). Evaluation of ICA-AROMA and alternative strategies for motion artifact removal in resting state fMRI. Neuroimage, 112, 278-287. Shirer, W. R., Ryali, S., Rykhlevskaia, E., Menon, V., & Greicius, M. D. (2012). Decoding subject-driven cognitive states with whole-brain connectivity patterns. Cerebral cortex, 22(1), 158-165. Snyder, A., Ma, L., Steinberg, J., Woisard, K., & Moeller, F. G. (2021). Dynamic causal modeling self-connectivity findings in the functional magnetic resonance imaging neuropsychiatric literature. Frontiers in Neuroscience, 15, 800. Sridharan, D., Levitin, D. J., & Menon, V. (2008). A critical role for the right fronto-insular cortex in switching between central-executive and default-mode networks. Proceedings of the National Academy of Sciences, 105(34), 12569-12574. Stephan, K. E., Penny, W. D., Daunizeau, J., Moran, R. J., & Friston, K. J. (2009). Bayesian model selection for group studies. Neuroimage, 46(4), 1004-1017. Woisard, K., Steinberg, J. L., Ma, L., Zuniga, E., Ramey, T., Lennon, M., Keyser-Marcus, L., and Moeller, F. G. (2021). Preliminary Findings of Weaker Executive Control Network Resting State fMRI Functional Connectivity in Opioid Use Disorder compared to Healthy Controls. Journal of Addiction Research & Therapy, 12(10). Woisard, K. (2022, March 7). Resting State fMRI Functional Connectivity in Cocaine Dependance. Retrieved from https://osf.io/t723q/ Woolrich, M. W., Behrens, T. E. J., Beckmann, C. F., Jenkinson, M., & Smith, S. M. (2004). Multilevel linear modelling for FMRI group analysis using Bayesian inference. NeuroImage, 21(4), 1732–1747.