A Novel Approach for Manual Segmentation of the Amygdala and Hippocampus in Neonate MRI

Abstract

The gross anatomy of the infant brain at term is fairly similar to that of the adult brain, but structures are immature, and the brain undergoes rapid growth during the first 2 years of life. Neonate magnetic resonance (MR) images have different contrasts compared to adult images, and automated segmentation of brain magnetic resonance imaging (MRI) can thus be considered challenging as less software options are available. Despite this, most anatomical regions are identifiable and thus amenable to manual segmentation. In the current study, we developed a protocol for segmenting the amygdala and hippocampus in T2-weighted neonatal MR images. The participants were 31 healthy infants between 2 and 5 weeks of age. Intra-rater reliability was measured in 12 randomly selected MR images, where 6 MR images were segmented at 1-month intervals between the delineations, and another 6 MR images at 6-month intervals. The protocol was also tested by two independent raters in 20 randomly selected T2-weighted images, and finally with T1 images. Intraclass correlation coefficient (ICC) and Dice similarity coefficient (DSC) for intra-rater, inter-rater, and T1 vs. T2 comparisons were computed. Moreover, manual segmentations were compared to automated segmentations performed by iBEAT toolbox in 10 T2-weighted MR images. The intra-rater reliability was high ICC ≥ 0.91, DSC ≥ 0.89, the inter-rater reliabilities were satisfactory ICC ≥ 0.90, DSC ≥ 0.75 for hippocampus and DSC ≥ 0.52 for amygdalae. Segmentations for T1 vs. T2-weighted images showed high consistency ICC ≥ 0.90, DSC ≥ 0.74. The manual and iBEAT segmentations showed no agreement, DSC ≥ 0.39. In conclusion, there is a clear need to improve and develop the procedures for automated segmentation of infant brain MR images.

Newborn amygdalar volumes are associated with maternal prenatal psychological distress in a sex-dependent way

“…Then, within each cluster, the sum-squared distance from each subject to each other subject was computed, and the subject with the minimum sum-squared distance was taken as the central-most subject of the cluster. The dual-contrast template constructed in the previous step was then warped to these 21 representative subjects, and provided for manual segmentation (without those doing the segmentation being aware that these were, in fact, 21 different versions of the same template) with standard procedures ( Hashempour et al, 2019 ).…”

“…Initially one template was segmented by the primary rater NH and senior rater JJT and externally reviewed by JDL. Then the other templates were segmented with standard procedures ( Hashempour et al, 2019 ). The final segmentations were a consensus between the primary and senior raters and were used in the subsequent labelling step.…”

Newborn left amygdala volume associates with attention disengagement from fearful faces at eight months

“…The structures of interest, i.e. the amygdalae, were manually labelled based on a previously described approach ( Hashempour N et al, 2019 ) on the dual-contrast templates – noting that 0.5 mm 3 resolution allowed more accurate segmentation. To ensure that these labels were accurate, we produced 21 variants of the template (each a non-linear transformation of the template to overlay one of the subjects in the population).…”

Maternal Pregnancy-Related Anxiety Is Associated With Sexually Dimorphic Alterations in Amygdala Volume in 4-Year-Old Children

“…The native anatomical images were preprocessed and segmented applying the volBrain pipeline (Manjón and Coupé, 2016). The segmentation of the amygdala volumes was amended by two raters according to the segmentation protocol by Hashempour et al (n.d.) and by use of the software ITK-SNAP (version 3.6.0) 3 (Yushkevich et al, 2006) and MNI Display 4 , respectively {interrater reliability [ICC(2,1)](Koo and Li, 2016): ICC (right amygdala) = 0.93, ICC (left amygdala) = 0.94}. Volumes are reported both uncorrected as well as corrected for total intracranial volume (TIV; assessed by use of volBrain).…”