Understanding the development of the human brain is challenging because imaging an immature brain encounters several difficulties. First, partial vol ume effects due to the small size of the brain asso ciated with an already complex pattern of gyrification (Fi g. 1) hamper cortex edges detection. Second, the GM WM contrast is weak due to unmyelinated white matte r. Finally, the human brain undergoes big and fast changes during the first months of post-natal life (e.g., the cranial perimeter increases by 0.5cm per week). However, these changes are not homogeneous a cro s the brain, some areas showing intense myelination and synaptogenesis (e.g., visual and mo tor areas) while others have a more protracted development (e.g., frontal areas) [1]. This maturat ion inhomogeneity produces important variation in tissue intensity on T2 Magnetic Resonance (MR) imag es. As shown in Figure 1, WM is much darker in myelinated areas than in non-myelinated ones. These specific characteristics of the infant brain explain why segmentation methods designed for T1 MR images of the adult brain are not optimal. To ou r knowledge, none of the well-known brain softwares, such as FreeSurfer, Caret, BrainSuite an d BrainVisa, have produced as yet automatic reconstruction of the infant’s cortical surface. As for T2-specific methods, atlas-based segmentatio s of the cortex [2, 3] require accurate and robust brain templates. However, variations of tissue cont rast are rapid and asynchronous during the first mo nths of life. We have not used brain atlas because we be li ve that multiple atlases would be required to ca pture the anatomical variability of infant brains. Xue and colleagues [4] have recently developed the only, to our knowledge, atlas-free automatic method for use in preterms and newborns. Their appr oach is mainly based on local estimates of tissue intensity to deal with intensity fluctuations acros s the brain. However, the GM-WM contrast quickly decreases in areas being myelinated during the firs t months of life. Thus, the GM-WM interface would not be properly detected in those regions using tis sue intensity alone. Here we propose features to ameliorate this problem.
[1]
P. Yakovlev,et al.
The myelogenetic cycles of regional maturation of the brain
,
1967
.
[2]
E. Duchesnay,et al.
A framework to study the cortical folding patterns
,
2004,
NeuroImage.
[3]
John H. Gilmore,et al.
Automatic segmentation of MR images of the developing newborn brain
,
2005,
Medical Image Anal..
[4]
Simon K. Warfield,et al.
Highly Accurate Segmentation of Brain Tissue and Subcortical Gray Matter from Newborn MRI
,
2006,
MICCAI.
[5]
Daniel Rueckert,et al.
Automatic segmentation and reconstruction of the cortex from neonatal MRI
,
2007,
NeuroImage.