Efficient Claustrum Segmentation in T2-weighted Neonatal Brain MRI Using Transfer Learning from Adult Scans

Purpose Intrauterine claustrum and subplate neuron development have been suggested to overlap. As premature birth typically impairs subplate neuron development, neonatal claustrum might indicate a specific prematurity impact; however, claustrum identification usually relies on expert knowledge due t...

Full description

Saved in:
Bibliographic Details
Published in:Clinical neuroradiology (Munich) 2022-09, Vol.32 (3), p.665-676
Main Authors: Neubauer, Antonia, Li, Hongwei Bran, Wendt, Jil, Schmitz-Koep, Benita, Menegaux, Aurore, Schinz, David, Menze, Bjoern, Zimmer, Claus, Sorg, Christian, Hedderich, Dennis M.
Format: Article
Language:English
Subjects:
Adults
Age
Algorithms
Automation
Brain research
Datasets
Deep learning
Failure analysis
Infants (Premature)
Investigations
Magnetic resonance imaging
Medicine
Medicine & Public Health
Neural networks
Neurology
Neurons
Neurophysiology
Neuroradiology
Neurosurgery
Original
Original Article
Premature birth
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Purpose Intrauterine claustrum and subplate neuron development have been suggested to overlap. As premature birth typically impairs subplate neuron development, neonatal claustrum might indicate a specific prematurity impact; however, claustrum identification usually relies on expert knowledge due to its intricate structure. We established automated claustrum segmentation in newborns. Methods We applied a deep learning-based algorithm for segmenting the claustrum in 558 T2-weighted neonatal brain MRI of the developing Human Connectome Project (dHCP) with transfer learning from claustrum segmentation in T1-weighted scans of adults. The model was trained and evaluated on 30 manual bilateral claustrum annotations in neonates. Results With only 20 annotated scans, the model yielded median volumetric similarity, robust Hausdorff distance and Dice score of 95.9%, 1.12 mm and 80.0%, respectively, representing an excellent agreement between the automatic and manual segmentations. In comparison with interrater reliability, the model achieved significantly superior volumetric similarity ( p  = 0.047) and Dice score ( p  
ISSN:1869-1439
1869-1447
DOI:10.1007/s00062-021-01137-8