Cross Data Set Generalization of Ultrasound Image Augmentation using Representation Learning: A Case Study

Current Directions in Biomedical Engineering

Hagenah Jannis¹  Wulff Daniel²  Mehdi Mohamad²  Ernst Floris² 

[1] Computational Health Informatics Lab, Institute of Biomedical Engineering, University of Oxford, Old Road Campus Research Building,OxfordOX3 7DQ, UK;Institute for Robotics and Cognitive Systems, University of Luebeck, Ratzeburger Allee 160, 23562Luebeck, Germany;

关键词: 2d ultrasound;    variational autoencoder;    generative adversarial network;    latent space;   

DOI  :  10.1515/cdbme-2021-2193

来源: DOAJ

【 摘 要 】

Data augmentation is a common method to make deep learning assessible on limited data sets. However, classical image augmentation methods result in highly unrealistic images on ultrasound data. Another approach is to utilize learning-based augmentation methods, e.g. based on variational autoencoders or generative adversarial networks. However, a large amount of data is necessary to train these models, which is typically not available in scenarios where data augmentation is needed. One solution for this problem could be a transfer of augmentation models between different medical imaging data sets. In this work, we present a qualitative study of the cross data set generalization performance of different learning-based augmentation methods for ultrasound image data. We could show that knowledge transfer is possible in ultrasound image augmentation and that the augmentation partially results in semantically meaningful transfers of structures, e.g. vessels, across domains.

【 授权许可】

Unknown