Deep Learning Models May Spuriously Classify Covid-19 from X-ray Images Based on Confounders
This study highlights a critical issue for medical AI, demonstrating that deep learning models for Covid-19 diagnosis from X-rays can be misled by spurious correlations, potentially leading to unreliable clinical tools.
This paper investigates the robustness of deep learning models for Covid-19 detection from chest X-ray images and finds that high-performing models often rely on confounding features related to data source or image processing artifacts, rather than actual lung pathology. This suggests that the models may be making diagnoses based on factors like patient age or image artifacts.
Identifying who is infected with the Covid-19 virus is critical for controlling its spread. X-ray machines are widely available worldwide and can quickly provide images that can be used for diagnosis. A number of recent studies claim it may be possible to build highly accurate models, using deep learning, to detect Covid-19 from chest X-ray images. This paper explores the robustness and generalization ability of convolutional neural network models in diagnosing Covid-19 disease from frontal-view (AP/PA), raw chest X-ray images that were lung field cropped. Some concerning observations are made about high performing models that have learned to rely on confounding features related to the data source, rather than the patient's lung pathology, when differentiating between Covid-19 positive and negative labels. Specifically, these models likely made diagnoses based on confounding factors such as patient age or image processing artifacts, rather than medically relevant information.