Connectome-based fingerprinting of clinical and functional outcomes in veterans

点击翻译按钮获取中文摘要

It is increasingly recognized that returning Veterans and Service Members of Operation Enduring Freedom/Operation Iraqi Freedom/Operation New Dawn (OEF/OIF/OND) suffer from co-occurring psychological and physical conditions that impede reintegration and make efficient and effective treatment planning difficult, if not impossible. Though it is clear that several diagnoses, particularly mTBI, PTSD, and depression are prevalent in trauma-exposed veterans, we have recently shown that specific patterns of co- occurrence of these disorders are key to understanding their functional consequences. The proposed research will investigate whether these patterns of clinical co-occurrence hold the key to discovering their neurobiological consequences. Specifically, this proposal will develop an innovative set of neuroimaging methods to determine whether these co-occurring disorders have specific neural fingerprints. Identifying such fingerprints would allow us to make diagnostic inferences about individual Veterans, and could help predict treatment outcomes and develop neurobiologically-informed interventions. To do this, the proposed studies will take advantage of existing data and will develop cutting-edge machine learning techniques and analytic procedures. Once developed, the proposed studies will poise the PI for multiple competitive Merit applications to apply these novel neural fingerprinting techniques in translational research. DESIGN AND METHODS: The proposed studies use a general set of techniques at the forefront of an exciting new era for brain imaging- MRI-based “fingerprinting”, or measuring and modeling the reproducible and yet substantial individual variation in the fMRI-based connectome (functional connectivity). This approach has the potential to translate population-based studies to investigations of the individual patient and precision medicine approaches. In Veterans, a recent study by Georgopoulos (co-I) and colleagues was able to successfully diagnose PTSD in a small, homogenous sample without comorbidities, using fMRI-based neural fingerprinting. The goal of the proposed studies is to replicate and optimize this work, as well as determine the feasibility of extracting unique and reliable neural fingerprints for veterans with complex co-occurring conditions (comorbid PTSD, mTBI, and depression). OBJECTIVES. Aim 1: Determine the feasibility of resting fMRI connectivity to predict PTSD in a polymorbid sample. Hypothesis: MRI-based fingerprinting will successfully diagnose PTSD above chance demonstrating the feasibility and validity of this fingerprinting analysis. However, diagnostic accuracy will be substantially reduced in this polymorbid sample thus demonstrating the need for future work to more finely characterize neural fingerprints associated with deployment trauma. Aim 2: Optimize this neural fingerprinting of PTSD across six different brain parcellation methods for defining the fMRI connectome. Aim 3: Determine the preliminary ability for resting fMRI connectivity to predict the functionally-relevant deployment trauma phenotype (DTP; comorbid mTBI, PTSD & depression). These Aims will motivate one or more Merit proposals to use neural fingerprinting to characterize a range of deployment- related pathologies, predict future functional outcomes, as well as guide the development of novel interventions.