Cortical parcellation for neonatal brains

Abstract

In the absence of a neonatal template with cortical subregion labels, it can be extremely difficult to obtain cortical parcellation of new neonatal brain images automatically. This paper addresses this problem by utilizing adult templates with rich cortical annotation and a neonatal template with simple tissue labels. Theoretical feasibility is assured because of the preservation of brain putative cytoarchitectonic boundaries from birth to adulthood. We use large deformation registration to propagate neuroanatomical labels from adult to neonatal brain and perform multi-atlas labeling based on accurate prior-based tissue segmentation. We evaluate the repeatability of the labeling by cross-validation with training and testing data. Preliminary results show interesting relationship between the volumes of labels and independent measures of neonatal development and maternal characteristics.

A review on automatic fetal and neonatal brain MRI segmentation

“…INU correction can be performed prior to the segmentation and/or inherently in the segmentation process. The N3 (Sled et al, 1998) and N4 algorithms (Tustison et al, 2010) are commonly used to perform INU correction in the perinatal brain (Makropoulos et al, 2014;Wu et al, 2014;Tourbier et al, 2015;Wang et al, 2015;Rajchl et al, 2016;Serag et al, 2016). Filtering techniques such as anisotropic diffusion have been further employed in the literature to reduce noise in the images while preserving the edges (Prastawa et al, 2005;Weisenfeld and Warfield, 2009;Gui et al, 2012b).…”

“…However, due to limited manual delination, quantitative evaluation on early preterm brain is presented for a few brain structures. Wu et al (2014) propagated labels from 20 adult OASIS templates. After an initial tissue segmentation based on EM, they parcellate the cortical ribbon into 62 regions with the label fusion method proposed by Wang et al (2012a).…”

Accurate Learning with Few Atlases (ALFA): an algorithm for MRI neonatal brain extraction and comparison with 11 publicly available methods

“…Whole-brain segmentation, also known as brain extraction or skull stripping, is the process of segmenting an MR image into brain and non-brain tissues. It is the first step in most neuroimage pipelines including: brain tissue segmentation and volumetric measurement 8 9 10 11 12 ; template construction 13 14 15 ; longitudinal analysis 16 17 18 19 ; and cortical and sub-cortical surface analysis 20 21 22 23 . Accurate brain extraction is critical because under- or over-estimation of brain tissue voxels cannot be salvaged in successive processing steps, which may lead to propagation of error through subsequent analyses.…”

Parcellation of the Healthy Neonatal Brain into 107 Regions Using Atlas Propagation through Intermediate Time Points in Childhood

“…For example, the UNC atlas was created using image registration and label fusion to propagate an adult brain atlas to 95 neonates through 2 and 1 year old templates (Tzourio-Mazoyer et al, 2002 ; Shi et al, 2011 ). Wu and colleagues used large deformation registration to propagate 62 neuroanatomical labels from adults to 15 neonatal brains and performed multi-atlas labeling based on accurate prior-based tissue segmentation (Wu et al, 2014 ). Makropoulos and colleagues performed multi-atlas segmentation by label fusion using the ALBERTs atlas (Makropoulos et al, 2014 ), and subsequently propagated the segmentations (plus labels of cortical ribbon) to the coordinate space of Serag et al ( 2012a ) and averaged these data with an age kernel at each timepoint to create a 4D atlas with 87 labeled structures (Makropoulos et al, 2016 ).…”