An optimised tract-based spatial statistics protocol for neonates: Applications to prematurity and chronic lung disease

Abstract

Preterm birth is associated with altered white matter microstructure, defined by metrics derived from diffusion tensor imaging (DTI). Tract-based spatial statistics (TBSS) is a useful tool for investigating developing white matter using DTI, but standard TBSS protocols have limitations for neonatal studies. We describe an optimised TBSS protocol for neonatal DTI data, in which registration errors are reduced. As chronic lung disease (CLD) is an independent risk factor for abnormal white matter development, we investigate the effect of this condition on white matter anisotropy and diffusivity using the optimised protocol in a proof of principle experiment.

DTI data were acquired from 93 preterm infants (48 male) with a median gestational age at birth of 28^+ 5 (23^+ 4–35^+ 2) weeks at a median postmenstrual age at scan of 41^+ 4 (38^+ 1–46^+ 6) weeks. Nineteen infants developed CLD, defined as requiring supplemental oxygen at 36 weeks postmenstrual age. TBSS was modified to include an initial low degrees-of-freedom linear registration step and a second registration to a population-average FA map.

The additional registration steps reduced global misalignment between neonatal fractional anisotropy (FA) maps. Infants with CLD had significantly increased radial diffusivity (RD) and significantly reduced FA within the centrum semiovale, corpus callosum and inferior longitudinal fasciculus (p < 0.05) compared to their peers, controlling for degree of prematurity and age at scan.

The optimised TBSS protocol improved reliability for neonatal DTI analysis. These data suggest that potentially modifiable respiratory morbidity is associated with widespread altered white matter microstructure in preterm infants at term-equivalent age.

Introduction

The incidence of preterm birth is increasing, and it is associated with a high prevalence of neurocognitive impairment in childhood (Delobel-Ayoub et al., 2009, Marlow et al., 2005) and adverse functional consequences that appear to persist into adolescence and early adulthood (Aarnoudse-Moens et al., 2009, Hille et al., 2007, Rushe et al., 2001). The perinatal factors that underlie abnormal brain development are not fully understood; however, respiratory illness is implicated because CLD is a risk factor for neurodevelopmental impairment independent of the degree of prematurity at birth (Hansen et al., 2004, Short et al., 2003, Westby Wold et al., 2009).

Diffusion tensor imaging (DTI) can provide voxelwise, scalar measures of water diffusion within the brain that reflect the underlying tissue microstructure (Basser et al., 1994, Basser and Pierpaoli, 1996). Low water diffusivity and high fractional anisotropy are characteristic of well-developed, myelinated white matter tracts, and are altered in experimental models of dysmyelination and neonatal white matter injury (Song et al., 2005, Wang et al., 2009). Compared with term-born controls, decreases in fractional anisotropy and increases in diffusivity have been observed within the white matter of preterm infants (Cheong et al., 2009, Counsell et al., 2006, Huppi et al., 1998, Miller et al., 2002, Partridge et al., 2004). These alterations are dependent on the degree of prematurity at birth (Anjari et al., 2007) and closely linked to neurodevelopmental outcome at 2 years corrected age (Counsell et al., 2008, Krishnan et al., 2007).

Most analyses of neonatal DTI have employed regions of interest (ROI) selected a priori to measure diffusion parameters directly in native diffusion tensor images (Cheong et al., 2009, Counsell et al., 2006, Kaukola et al., 2009, Partridge et al., 2004). Although this technique avoids computationally demanding image registration and offers a valid method to infer groupwise variation in white matter development, it remains subjective and manually intensive and does not easily allow for comparisons across many subjects. Tract-based spatial statistics (TBSS) is a powerful tool for studies of the adult brain that improves the sensitivity, objectivity and interpretability of multi-subject DTI analysis (Smith et al., 2006). TBSS achieves this through carefully tuned alignment of fractional anisotropy maps to a standard-space template, followed by projection of individual data onto a skeletonised representation of major white matter tracts common to the group for voxelwise analysis.

Neonatal DTI data are of lower resolution and contrast compared to adult data, and these differences together with relatively wide variations in brain size and complexity within neonatal populations pose technical challenges to DTI processing tools including TBSS. In previous work we have overcome some of these challenges by using a representative, individual FA map as a study-specific template (Anjari et al., 2007, Anjari et al., 2009); however, registration between neonatal FA maps can still fail (Bassi et al., 2008). In addition, registration within TBSS has been empirically optimized for accurate alignment to a group-average map that is distinctly different in both contrast and smoothness to an individual FA map.

The primary aim of this study was to improve the reliability of TBSS in neonates by implementing two modifications: the inclusion of an initial low degrees-of-freedom linear registration to improve global alignment between neonatal FA maps and a second registration to a population-average FA map to produce accurate projection of individual data on a skeleton for subsequent multi-subject analysis of white matter diffusivity and anisotropy.

As a secondary aim, using the optimised TBSS protocol, we investigated the association between prematurity, CLD and white matter microstructure in a cohort of preterm infants imaged at term-equivalent age.