Convolutional neural networks (CNNs) have been widely and successfully used for medical image segmentation. However, CNNs are typically considered to require large numbers of dedicated expert-segmented training volumes, which may be limiting in practice. This work investigates whether clinically obtained segmentations which are readily available in picture archiving and communication systems (PACS) could provide a possible source of data to train a CNN for segmentation of organs-at-risk (OARs) in radiotherapy treatment planning. In such data, delineations of structures deemed irrelevant to the target clinical use may be lacking. To overcome this issue, we use multi-label instead of multi-class segmentation. We empirically assess how many clinical delineations would be sufficient to train a CNN for the segmentation of OARs and find that increasing the training set size beyond a limited number of images leads to sharply diminishing returns. Moreover, we find that by using multi-label segmentation, missing structures in the reference standard do not have a negative effect on overall segmentation accuracy. These results indicate that segmentations obtained in a clinical workflow can be used to train an accurate OAR segmentation model.
@article{arxiv.1911.04967,
title = {Exploiting Clinically Available Delineations for CNN-based Segmentation in Radiotherapy Treatment Planning},
author = {Louis D. van Harten and Jelmer M. Wolterink and Joost J. C. Verhoeff and Ivana Išgum},
journal= {arXiv preprint arXiv:1911.04967},
year = {2019}
}