Continuing the follow-up of paper reviews on issues of interest for The Information Age, I came across today with a paper which combines several topics that are currently in the ranks of cutting edge innovative technologies. The fields are Convolutional Neural Networks (CNNs), Artificial Neural Networks and imaging in Neuroscience studies.
Whilst Convolutional Neural Networks and Artificial Neural Networks aren’t new technology, what is happening nowadays with those subjects is interesting enough that there is already a distinction between the traditional approach and the more innovative ones, with good reasons, given the pace of new developments that appear on an almost daily basis.
Today’s paper belongs to the set of the new innovative developments within Convolutional Neural Networks with the introduction of a novel technology called BrainNetCNN, that improves on the traditional approach, by replacing image-based CNNs with novel edge-to-edge, edge-to-node and node-to-graph convolutional filters that leverage the topological locality of structural brain networks. The full abstract conveys the bigger picture:
We propose BrainNetCNN, a convolutional neural network (CNN) framework to predict clinical neurodevelopmental outcomes from brain networks. In contrast to the spatially local convolutions done in traditional image-based CNNs, our BrainNetCNN is composed of novel edge-to-edge, edge-to-node and node-to-graph convolutional filters that leverage the topological locality of structural brain networks. We apply the BrainNetCNN framework to predict cognitive and motor developmental outcome scores from structural brain networks of infants born preterm. Diffusion tensor images (DTI) of preterm infants, acquired between 27 and 46 weeks gestational age, were used to construct a dataset of structural brain connectivity networks. We first demonstrate the predictive capabilities of BrainNetCNN on synthetic phantom networks with simulated injury patterns and added noise. BrainNetCNN outperforms a fully connected neural-network with the same number of model parameters on both phantoms with focal and diffuse injury patterns. We then apply our method to the task of joint prediction of Bayley-III cognitive and motor scores, assessed at 18 months of age, adjusted for prematurity. We show that our BrainNetCNN framework outperforms a variety of other methods on the same data. Furthermore, BrainNetCNN is able to identify an infant’s post menstrual age to within about 2 weeks. Finally, we explore the high-level features learned by BrainNetCNN by visualizing the importance of each connection in the brain with respect to predicting the outcome scores. These findings are then discussed in the context of the anatomy and function of the developing preterm infant brain.