Development of quantitative MRI DTI analysis tool for preterm neonate

Johns Hopkins University, Baltimore MD

Investigators

Linked publications & trials

Abstract

DESCRIPTION (provided by applicant): We will establish a quantification method for neonatal brain MRI to evaluate the abnormalities of preterm born neonates. In the U.S., approximately 12% of all neonates are born preterm (<37 weeks gestation), with 2% of these being very preterm (VPB, 28-32 weeks gestation). The percentage of preterm births has been increasing over the last ten years, partly due to improved neonatal care for preterm infants. Nevertheless, about half of the VPB infants may develop clinically-evident neurological or psychological disorders and the number could be even higher if subtle functional abnormalities are included. The extent of neurocognitive deficits in the late preterm infants (33-36 weeks gestation) is also unknown. However, most of the neuro- cognitive deficits are not easily detected during the first year of life. To benefit from early intervention and to develop more deficits-specific interventions, we need methods to detect and quantify brain abnormalities at an early stage. MRI is one of the most promising and sensitive methods to detect subtle anatomic abnormalities in the neonatal brain. Previous brain MRI studies have found some correlations between several types of abnormalities and neurological outcomes, but there are also reports that found no relationship between signal alterations and neurological outcomes. Hence, the current knowledge does not justify the use of MRI for routine clinical evaluations. To optimize the usefulness of MRI for neonatal and pediatric care, systematic research, based on quantitative image analysis and functional correlation, is needed. The proposed method is based on two core technologies for the quantification of neonatal brain anatomy: a deformable brain atlas with detailed anatomic information and a highly elastic topology-preserved warping method. The combination will provide multiple MR parameters from 176 automatically segmented brain structures. The goals of this project are to establish an atlas-based, automated quantification method for the neonatal brain, to evaluate the detail anatomy of premature neonates at a term-equivalent age. The overall hypothesis is that our DTI-guided quantitative brain analysis will sensitively detect anatomical abnormalities of preterm born neonates in region- specific manner. We have four specific aims. In Aim 1, we will create a multi-contrast (T1-, T2-weighted, and DTI) normal-term neonatal brain atlas for quantitative brain analysis, which will be a statistical representation of the population (Bayesian atlas). In Aim 2, we will combine the Bayesian atlas with highly elastic topology- preserved warping (Large Deformation Diffeomorphic Metric Mapping, LDDMM) for automated brain segmentation and test the segmentation accuracy. In Aim 3, we will use the combination of the Bayesian atlas and LDDMM to perform T1/T2/DTI quantification of term neonatal brain MRIs. In Aim 4, we will apply the method to the brain MRIs from term-equivalent preterm born infants (born at 28-36 weeks gestational age) and compare the MR parameters to those in the term infants. This study will be a first step toward seeking very early prognostic indicators for functional outcomes of the anatomical brain abnormalities in preterm births.