Semantic segmentation of computed tomography for radiotherapy with deep learning: Compensating insufficient annotation quality using contour augmentation

In radiotherapy treatment planning, manual annotation of organs-at-risk and target volumes is a difficult and time-consuming task, prone to intra and inter-observer variabilities. Deep learning networks (DLNs) are gaining worldwide attention to automate such annotative tasks because of their ability to capture data hierarchy. However, for better performance DLNs require large number of data samples whereas annotated medical data is scarce. To remedy this, data augmentation is used to increase the training data for DLNs that enables robust learning by incorporating spatial/translational invariance into the training phase. Importantly, performance of DLNs is highly dependent on the ground truth (GT) quality: if manual annotation is not accurate enough, the network cannot learn better than the annotated example. This highlights the need to compensate for possibly insufficient GT quality using augmentation, i.e., by providing more GTs per image, in order to improve performance of DLNs. In this work, small random alterations were applied to GT and each altered GT was considered as an additional annotation. Contour augmentation was used to train a dilated U-Net in multiple GTs per image setting, which was tested on a pelvic CT dataset acquired from 67 patients to segment bladder and rectum in a multi-class segmentation setting. By using contour augmentation (coupled with data augmentation), the network learnt better than with data augmentation only, as it was able to correct slightly offset contours in GT. The segmentation results produced were quantified using spatial overlap, distance-based and probabilistic measures. The Dice score for bladder and rectum are reported as 0.88 +/- 0.19 and 0.89 +/- 0.04, whereas the average symmetric surface distance are 0.22 +/- 0.09 mm and 0.09 +/- 0.05 mm, respectively.