Automatic coronavirus disease 2019 diagnosis based on chest radiography and deep learning – Success story or dataset bias?

Abstract

Results in this study confirm that when trained on a combinatory dataset, CNNs tend to learn the origin of the CXRs rather than the presence or absence of disease, a behavior known as short-cut learning. The bias is shown to originate from differences in overall pixel values rather than embedded text or symbols, despite consistent image pre-processing. When trained on a reliable, and realistic single-source dataset in which non-lung pixels have been masked, CNNs currently show limited sensitivity (<70%) for COVID-19 infection in CXR, questioning their use as a reliable automatic screening tool.

Ähnliche Arbeiten