Mapping Subject-Specific Structural and Functional Connectivity to Parse the Unique Contributions of Subconcussive Blast, Mild TBI, and PTSD

点击翻译按钮获取中文摘要

ABSTRACT Subconcussive blast is, by definition, not associated with acute or chronic phase symptoms. However, neuroimaging offers initial evidence of damage to white matter following subconcussive blast. As with mTBI, subconcussive blast is associated with tremendous spatial and inter-individual heterogeneity resulting from variability in the events surrounding blast. A significant limitation of subconcussive, mTBI, and PTSD neuroimaging research results from the comparison of group-averaged data of affected individuals to matched controls that tends to mask individual-specific features. Fortunately, recent methodological advances in the acquisition and analysis of resting-state fMRI (rs-fMRI) now offer an unprecedented ability to map individual- subject-level functional connectivity. Long rs-fMRI scans (> 25 min) coupled with advanced individual-specific parcellation methods provides quantification of inter-subject heterogeneity. 210 veterans from 4 subject categories: (1) subconcussive (2) concussive mTBI (3) unexposed controls and (4) blast-unexposed PTSD, will undergo (1) long rs-fMRI scans, structural MRI, diffusion MRI, (2) assessment of blast exposure, head impact, and associated clinical symptoms, (3) psychiatric symptoms, and (4) a neurocognitive battery. In Specific Aim 1, we will compare individual-specific maps of functional and structural networks in every subject and between the 4 subject categories. We hypothesize unique patterns of functional and structural connectivity among individual subjects in each category. In Specific Aim 2, we will measure the congruence between structural and functional network architectures. We hypothesize that the structure-function uncoupling in subconcussive and mTBI groups will originate from structural alterations but the source of uncoupling in PTSD will be from functional alterations. In Specific Aim 3, we will relate functional and structural network connectivity to the events surrounding blast exposure, mTBI symptoms, PTSD symptoms, PTSD symptom clusters, depression symptoms, other neuropsychiatric symptoms, and neurocognitive performance. Support vector regression methods will train algorithms capable of recognizing individual-specific features of functional connectivity, structural connectivity, and structure-function coupling that correspond to the subconcussive, mTBI, PTSD, and control groups. An overwhelming challenge in the diagnosis and clinical management of veterans, is delineating a complex array or overlapping exposures, histories, symptoms, and clinical signs into distinct etiologically defined diagnostic entities. We propose analytic methods for unraveling heterogeneity, which index brain network connectivity metrics derived from neuroimaging. We are taking important strides toward discovering knowledge that is essential to future development of effective treatments and clinical management. Mapping individual-specific etiologic pathways and mechanisms of disease will move us one step closer to implementing precision medicine for Veterans’ health.