CARE4Kids: Imaging Biomarker Core

点击翻译按钮获取中文摘要

PROJECT SUMMARY/ABSTRACT – Imaging Biomarker Core The CARE4Kids Center Without Walls Project aims to discover measurable biomarkers of persistent post- concussive symptoms (PPCS) and use these markers to characterize subgroups (endophenotypes) that will inform prognosis and potential treatment. Advanced neuroimaging offers great promise in revealing how concussion affects brain structure and function, but relatively small sample sizes have limited the reliability and generalizability of imaging studies to date. The overall goal of the Imaging Biomarker Core is to collect state-of- the-art multimodal brain imaging data that is maximally informative about outcomes in children and adolescents with concussion. Multimodal imaging will be collected in the Development Cohort, at baseline only, and used to develop and test models of outcome prediction that will be examined in the Validation Cohort. The neuroimaging protocol is designed to maximize data quality and minimize discomfort for our injured, younger patient population and is based on existing multi-site efforts (i.e., ABCD, UK Biobank) and NINDS imaging recommendations. The Imaging Biomarker Core has the following Specific Aims: Specific Aim 1: Discover which multimodal brain imaging measures best predict clinical outcomes in concussion (separately, and when combined with other biomarkers and clinical measures). Specific Aim 2: Neuroimaging site qualification and training – site qualification using pilot data and coordinated training will prepare sites to collect high-quality data. Specific Aim 3: Collect high-quality multimodal brain MRI – 360 participants will be scanned across 6 sites in the Development Cohort. Specific Aim 4: Ongoing data quality control – ongoing quality control through phantoms, volunteers, and data review will ensure high-quality data that is maximally comparable across sites. Specific Aim 5: Analyze Data: Transfer images and derived measures to the University of Utah Data Coordinating Center (U-DCC) to disseminate to the research community. We will use standardized, validated, publicly-available protocols to process the neuroimaging data. Measures of regional brain volumes, cortical geometry, white matter organization, functional connectivity, perfusion, and neuropathology (i.e., microbleeds and white matter hyperintensities) will be compared between groups and examined for correlation with outcome measures. In addition to these simpler approaches, working with the U-DCC, we will employ machine learning approaches to identify which combination of imaging, demographic, and clinical measures best predicts functional outcome.