Bayesian Generative Models for Knowledge Transfer in MRI Semantic Segmentation Problems
This addresses the challenge of small datasets in medical MRI segmentation, which is a domain-specific incremental improvement.
The paper tackled the problem of applying deep learning segmentation to small medical datasets by proposing a knowledge transfer method using a Generative Bayesian Prior network, achieving the best Dice Similarity Coefficient results on small subsets of the BRATS2018 database.
Automatic segmentation methods based on deep learning have recently demonstrated state-of-the-art performance, outperforming the ordinary methods. Nevertheless, these methods are inapplicable for small datasets, which are very common in medical problems. To this end, we propose a knowledge transfer method between diseases via the Generative Bayesian Prior network. Our approach is compared to a pre-train approach and random initialization and obtains the best results in terms of Dice Similarity Coefficient metric for the small subsets of the Brain Tumor Segmentation 2018 database (BRATS2018).