Classification of Gulf War Illness Patients vs Control Veterans Using fMRI Functional Connectivity

Submission No:

2493 

Submission Type:

Abstract Submission 

Authors:

Unal Sakoglu1, Mounika Galla1, Sasanka Bhamidipati1, Kaundinya Gopinath2, Bruce Crosson2, Robert Haley3

Institutions:

1University of Houston - Clear Lake, Houston, TX, 2Emory University, Atlanta, GA, 3University of Texas Southwestern Medical Center at Dallas, Dallas, TX

E-Poster

Introduction:

Gulf War Illness (GWI) is a medical condition characterized by multiple symptoms which indicate brain function deficits in emotion, cognitive, pain and somatosensory domains [1-3]. It affects about 200,000 of the 1991 Gulf War veterans. Prior neuroimaging studies confirm presence of structural, functional and metabolic brain impairments in GWI [4-6]; however, GWI is still poorly understood. Recently, functional neuroimaging technology, especially fMRI, has improved tremendously, with recent attention towards resting-state functional connectivity (FC) and independent component analysis (ICA) of the brain [7-9]. ICA can reveal which parts of the brain act as (statistically) independent functional networks [7]. FC analysis, based on ICA, can provide insight into overall functional connectivity between networks; differences in the FC of the brain networks under different brain conditions can be studied and used as features for classification [9]. In this study we used ICA-based FC as features to classify the GWI patients vs. normal control (NC) veterans.

Methods:

23 GWI veterans (mean age 49.4 yrs.) and 30 NC (mean 49.8 yrs.) were scanned in a Siemens 3T Tim Trio MRI scanner using a 12-channel head coil. Written informed consent was obtained from all participants in the protocol approved by the local IRB. Whole brain resting-state fMRI (rsfMRI) data were acquired with a 10-min whole-brain gradient echo EPI (TR/TE/FA=2000/24ms/90°, 3mm×3mm×3.5mm resolution). RsfMRI preprocessing steps included attenuation of signal related to subject-motion and physiological responses, advanced ICA-based artifact reduction techniques, spatial smoothing with FWHM=6mm isotropic Gaussian kernel. Group ICA [10] were performed separately for each group, the number of independent components (ICs) were determined (21 ICs for NC and 23 ICs for patients) using MDL criteria [7]. Subsequently, an overall combined group ICA was done with the total number of ICs (44) [11]. After removal of 5 ICs which correspond to artifacts, the remaining 39 ICs were paired to obtain 39*38/2=741 FC values, which were used as initial features. The features were ranked and reduced to only 4 features (FC pairs) by using the Wrapper subset evaluator with "best-first" forward-search utility in Weka machine learning toolbox [12]. Classification was performed using the reduced number of features using Weka.

Results:

Combined group ICA, when unrestricted, resulted in 22 ICs using the MDL. Therefore, ICA with 44 ICs represents 2x the ICs. Five artifactual ICs were were removed from the classification analyses. After feature reduction to 4 FCs as described in the Methods section, using 10-fold cross-validation, a Naive Bayes Classifier algorithm classified 28/30 of NC and 19/23 of the GWI correctly, which corresponds to sensitivity of 82.6%, specificity of 93.3%, and an overall classification accuracy of 88.7%. The 4 most discriminating FC pairs were the following, presented in Fig. 1:
FC1: IC2 (visual) & IC16 (motor);
FC2: IC12 (frontal) & IC43 (vis assoc ctx, retrosplen, post parahippo; post cing);
FC3: IC23 (bil Sylvian, s temp) & IC34 (L mid/sup temp);
FC4: IC24 (R lat fron) & IC33 (R par, pre SM, fron/par, post mid cing).
Mean FC values for the NC were higher than GWI for these FCs (p<0.05 for FC1 and FC4).
Supporting Image: Figure-01.png
   ·Figure 1. Functional connectivity between the 4 brain networks, which have the most discriminating power between the two groups, gulf war illness patients and normal control veterans.
 

Conclusions:

FC of multiple brain networks have good group discriminating power of 88.7%. FC1 is between higher visual processing areas and motor cortex; which is important for visuo-motor function, multisensory processing. FC2 could be related to top-down influence on visual functions. FC3 could be related to auditory and language functions. FC4 could be related to fronto-parietal and task-positive networks. Overall, these regions were previously shown to be involved in either visual, auditory, sensory-motor input processing, or semantic processing. Consistent with these findings, GWI veterans were reported to exhibit deficits in word-finding [13], in visual processing [14], and in fine motor skills [1].

Disorders of the Nervous System:

Traumatic Brain Injury
Disorders of the Nervous System Other

Imaging Methods:

BOLD fMRI

Modeling and Analysis Methods:

Classification and Predictive Modeling 2
fMRI Connectivity and Network Modeling 1

Keywords:

FUNCTIONAL MRI
Machine Learning

1|2Indicates the priority used for review