딥러닝 성능 향상을 위한 3D 의료 영상 증강법

Abstract

This paper proposes a 3D image augmentation method for improving the generalization performance of deep neural networks. It allows us to enrich the diversity of training data samples that is essential in medical image segmentation tasks, thus reducing the data overfitting problem caused by the fact the scale of medical image dataset is typically smaller. It also enables us to predict medical segmentation surfaces in Euclidean space without additional labeled datasets. This method includes image transformation functions, which are comprised of a spatial deformation and image intensity change, enabling the synthesis of complex effects such as variations in anatomy and image acquisition procedures. Our numerical experiments demonstrate that the proposed approach provides significant improvements over state-of-the-art methods for 3D medical image segmentation.