AUTHOR=Ding Yang , Acosta Rolando , Enguix Vicente , Suffren Sabrina , Ortmann Janosch , Luck David , Dolz Jose , Lodygensky Gregory A. 

TITLE=Using Deep Convolutional Neural Networks for Neonatal Brain Image Segmentation

JOURNAL=Frontiers in Neuroscience

VOLUME=Volume 14 - 2020

YEAR=2020

URL=https://www.frontiersin.org/journals/neuroscience/articles/10.3389/fnins.2020.00207

DOI=10.3389/fnins.2020.00207

ISSN=1662-453X

ABSTRACT=Deep learning neural networks are especially potent at dealing with structured data, such as images and volumes. Both modified LiviaNET and HyperDense-Net performed well at prior competition segmenting six-month old infant magnetic resonance images, but neonatal cerebral tissue type identification is challenging given its uniquely inverted tissue contrasts. Current study aims to evaluate the two architectures to segment neonatal brain tissues types at term equivalent age. Both networks were retrained over 24 pairs of neonatal T1 and T2 data from the developing human connectome project public data set and validated on another eight pairs against ground truth. We then reported the best performing model from training and its performance by computing Dice similarity coefficient (DSC) for each tissue type against eight test subjects. During the testing phase, single modality LiviaNET performed statistically significantly better when processing T2-weighted images than T1-weighted images across all tissue types, achieving mean DSC values of 0.903/0.904/0.884 for gray matter, white matter and cerebrospinal fluid respectively while requiring 30 hours of training and two minutes to segment each brain. HyperDense-Net achieved statistically significantly better test mean DSC values, obtaining 0.942/0.945/0.915 for the tissue types respectively and took 80 hours to train and eight minutes to segment. Our evaluation demonstrates that both neural networks can segment neonatal brains achieving previously reported performance. Both networks will be continuously retrained over increasingly larger repertoire of neonatal brain data and be made available through the Canadian Neonatal Brain Platform to better serve the neonatal brain imaging research community.