CT Image Synthesis Using Weakly Supervised Segmentation and Geometric Inter-Label Relations For COVID Image Analysis
This work addresses the challenge of data scarcity and variability in medical imaging for COVID-19 diagnosis, offering an incremental improvement over existing GAN-based synthesis methods by incorporating geometric inter-label relations.
The authors tackled the problem of limited and variable performance in medical image segmentation due to scarce pixelwise annotations and conventional data augmentation by proposing a GAN-based image synthesis method that learns geometric relationships between anatomical labels using weakly supervised segmentation. Their method improved COVID-19 lung CT segmentation, outperforming state-of-the-art methods on a public dataset.
While medical image segmentation is an important task for computer aided diagnosis, the high expertise requirement for pixelwise manual annotations makes it a challenging and time consuming task. Since conventional data augmentations do not fully represent the underlying distribution of the training set, the trained models have varying performance when tested on images captured from different sources. Most prior work on image synthesis for data augmentation ignore the interleaved geometric relationship between different anatomical labels. We propose improvements over previous GAN-based medical image synthesis methods by learning the relationship between different anatomical labels. We use a weakly supervised segmentation method to obtain pixel level semantic label map of images which is used learn the intrinsic relationship of geometry and shape across semantic labels. Latent space variable sampling results in diverse generated images from a base image and improves robustness. We use the synthetic images from our method to train networks for segmenting COVID-19 infected areas from lung CT images. The proposed method outperforms state-of-the-art segmentation methods on a public dataset. Ablation studies also demonstrate benefits of integrating geometry and diversity.