Exploiting Clinically Available Delineations for CNN-based Segmentation in Radiotherapy Treatment Planning

TL;DR

This study explores using readily available clinical delineations from PACS systems to train CNNs for organ-at-risk segmentation in radiotherapy, showing that even limited data can be effective with multi-label training.

Contribution

It demonstrates that clinical delineations, despite missing some structures, can effectively train CNNs for OAR segmentation using multi-label approaches.

Findings

Increasing training data beyond a small set yields diminishing returns.

Multi-label segmentation mitigates the impact of missing structures.

Clinical delineations can be sufficient for accurate CNN training.

Abstract

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.