Reducing Labelled Data Requirement for Pneumonia Segmentation using Image Augmentations
This addresses the bottleneck of costly data annotation for medical imaging, offering a domain-specific incremental improvement.
The study tackled the problem of reducing the need for expensive labeled data in pneumonia segmentation from chest X-rays by testing image augmentations, finding that rotate and mixup augmentations cut labeled data requirements by 70% while maintaining comparable performance in AUC and mean IoU.
Deep learning semantic segmentation algorithms can localise abnormalities or opacities from chest radiographs. However, the task of collecting and annotating training data is expensive and requires expertise which remains a bottleneck for algorithm performance. We investigate the effect of image augmentations on reducing the requirement of labelled data in the semantic segmentation of chest X-rays for pneumonia detection. We train fully convolutional network models on subsets of different sizes from the total training data. We apply a different image augmentation while training each model and compare it to the baseline trained on the entire dataset without augmentations. We find that rotate and mixup are the best augmentations amongst rotate, mixup, translate, gamma and horizontal flip, wherein they reduce the labelled data requirement by 70% while performing comparably to the baseline in terms of AUC and mean IoU in our experiments.