Label-Free Segmentation of COVID-19 Lesions in Lung CT
This addresses the challenge of data scarcity for automated COVID-19 diagnosis, offering a practical solution for medical imaging with incremental improvements in unsupervised segmentation.
The authors tackled the problem of segmenting COVID-19 lesions in lung CT scans without labeled data by developing a label-free method that uses pixel-level anomaly modeling from normal scans, achieving superior performance over unsupervised anomaly detection methods on three datasets.
Scarcity of annotated images hampers the building of automated solution for reliable COVID-19 diagnosis and evaluation from CT. To alleviate the burden of data annotation, we herein present a label-free approach for segmenting COVID-19 lesions in CT via pixel-level anomaly modeling that mines out the relevant knowledge from normal CT lung scans. Our modeling is inspired by the observation that the parts of tracheae and vessels, which lay in the high-intensity range where lesions belong to, exhibit strong patterns. To facilitate the learning of such patterns at a pixel level, we synthesize `lesions' using a set of surprisingly simple operations and insert the synthesized `lesions' into normal CT lung scans to form training pairs, from which we learn a normalcy-converting network (NormNet) that turns an 'abnormal' image back to normal. Our experiments on three different datasets validate the effectiveness of NormNet, which conspicuously outperforms a variety of unsupervised anomaly detection (UAD) methods.