Inherent Brain Segmentation Quality Control from Fully ConvNet Monte Carlo Sampling
This work provides automated quality control for brain segmentation, which is incremental but useful for clinical practice and large-scale data processing.
The paper tackled the problem of quality control in brain segmentation by introducing uncertainty metrics derived from a Bayesian fully convolutional neural network with Monte Carlo dropout sampling, showing that these metrics are highly correlated with segmentation accuracy and improve effect sizes in group analyses to be closer to manual annotations.
We introduce inherent measures for effective quality control of brain segmentation based on a Bayesian fully convolutional neural network, using model uncertainty. Monte Carlo samples from the posterior distribution are efficiently generated using dropout at test time. Based on these samples, we introduce next to a voxel-wise uncertainty map also three metrics for structure-wise uncertainty. We then incorporate these structure-wise uncertainty in group analyses as a measure of confidence in the observation. Our results show that the metrics are highly correlated to segmentation accuracy and therefore present an inherent measure of segmentation quality. Furthermore, group analysis with uncertainty results in effect sizes closer to that of manual annotations. The introduced uncertainty metrics can not only be very useful in translation to clinical practice but also provide automated quality control and group analyses in processing large data repositories.