Nonlinear microstructural changes in the right superior temporal sulcus and lateral occipitotemporal gyrus between 35 and 43 weeks in the preterm brain Alec Aeby a, b, ⁎, Patrick Van Bogaert a, b , Philippe David c , Danielle Balériaux c , Danièle Vermeylen d , Thierry Metens c , Xavier De Tiège b a Department of Pediatric Neurology, Université Libre de Bruxelles (ULB)-Hôpital Erasme, Brussels, Belgium b Laboratoire de Cartographie Fonctionnelle du Cerveau, ULB-Hôpital Erasme, Brussels, Belgium c Department of Neuroradiology, ULB-Hôpital Erasme, Brussels, Belgium d Department of Neonatology, ULB-Hôpital Erasme, Brussels, Belgium abstract article info Article history: Accepted 10 June 2012 Available online 17 June 2012 Using diffusion tensor imaging (DTI), we explored microstructural brain maturation in a population of 65 pre- term neonates who underwent magnetic resonance imaging between 35 and 43 weeks of corrected gestational age. A voxel-based analysis approach, statistical parametric mapping (SPM8), was used to evidence the nonlinear changes with the corrected gestational age in the regional distribution of mean diffusivity (MD), fractional an- isotropy (FA), longitudinal and transverse diffusivities (λ//and λ⊥). We found that FA changes nonlinearly with age around the right superior temporal sulcus and in the right lateral occipitotemporal gyrus, with FA de- crease between 34 and 39 weeks followed by FA increase from 40 weeks to 43 weeks. Considering the key role of these brain areas in verbal and non-verbal communicative behaviors, the effect of these microstructural changes in terms of early social network functional maturation needs to be assessed by joint functional and an- atomical studies. © 2012 Elsevier Inc. All rights reserved. Introduction The rapid and important developmental changes that occur in the brain during the third trimester of gestation (sulcation and changes in the organization and maturation of white matter) may be studied using magnetic resonance imaging (MRI). Although feasible (Zanin et al., 2011), fetal MRI still remains a technical challenge because of the many sources of artifacts, mainly related to fetal and maternal motions (Kasprian et al., 2008; Scifo et al., 2003). Studying brain de- velopment in the preterm is less susceptible to these artifacts and also of great interest (Berman et al., 2005; Glasel et al., 2011; Hüppi et al., 1998), as it provides insights into early brain development and how early birth may affect normal brain development (Mewes et al., 2006). Conventional brain MRI is used to explore myelination of the white matter (WM) (van der Knaap et al., 1990) as well as sulcation and gyrification of the cortex (Dubois et al., 2008a; Girard et al., 1995). Nevertheless, this technique is not sensitive enough to evalu- ate the microstructural correlates of brain development. Diffusion tensor imaging (DTI) is an MRI technique that assesses and quantifies water diffusion in biological tissues at a microstructural level. This technique takes advantage of the fact that, in the brain, water molecules diffuse more easily in the direction of the fibers than or- thogonally to study cortical and WM maturations. DTI indices like fractional anisotropy (FA), which expresses the fraction of the magni- tude of the diffusion tensor attributable to anisotropic diffusion, mean diffusivity (MD), which corresponds to the directionally averaged magnitude of water diffusion, and also the longitudinal and trans- verse diffusivities (λ//and λ ⊥), which express respectively the paral- lel and perpendicular diffusion of water molecules, are used to indirectly quantify brain microstructure (Berman et al., 2005; Ulug and van Zijl, 1999). Several studies have shown that in the WM between 34 and 40 weeks of gestational age (GA), λ ⊥ decreases faster than λ//, resulting in an increase in FA and a decrease in MD. Histological data suggest that these DTI modifications are related to regression of cytoplasmic arborization of immature oligodendrocytes followed by the progressive ensheatment of axons by myelinated and unmyelinated oligodendrocytes (Aeby et al., 2009; Berman et al., 2005; Zanin et al., 2011). In the cortex, FA was shown to decrease line- arly with age chiefly due to a preferential reduction in λ ⊥ (Deipolyi et al., 2005; McKinstry et al., 2002). Biologically, this might correspond to the arborization of basal dendrites from immature neurons, the in- nervation of the cortical plate by thalamocortical and corticocortical axons and early synaptogenesis (Marin-Padilla, 1988; Mrzljak et al., 1988). Up to now, most of the DTI studies have assumed that brain devel- opment and maturation is a linear process (Aeby et al., 2009; Berman et al., 2005; Deipolyi et al., 2005; Dubois et al., 2008b; Hüppi et al., NeuroImage 63 (2012) 104–110 ⁎ Corresponding author at: Department of Pediatric Neurology, ULB-Hôpital Erasme, 808 routes de Lennik, 1070 Brussels, Belgium. Fax: + 32 25558333. E-mail address: alec.aeby@erasme.ulb.ac.be (A. Aeby). 1053-8119/$ – see front matter © 2012 Elsevier Inc. All rights reserved. doi:10.1016/j.neuroimage.2012.06.013 Contents lists available at SciVerse ScienceDirect NeuroImage journal homepage: www.elsevier.com/locate/ynimg