article Open AccessTop 1% cited

Deep learning for chest radiograph diagnosis: A retrospective comparison of the CheXNeXt algorithm to practicing radiologists

PLoS Medicine · 2018 · Vol. 15(11) · pp. e1002686–e1002686

Pranav Rajpurkar✉(Stanford University)Jeremy Irvin(Stanford University)Robyn L. Ball(Stanford University)Kaylie Zhu(Stanford University)Brandon Yang(Stanford University)Hershel Mehta(Stanford University)Tony Duan(Stanford University)Daisy Yi Ding(Stanford University)Aarti Bagul(Stanford University)Curtis P. Langlotz(Stanford University)Bhavik N. Patel(Stanford University)Kristen W. Yeom(Stanford University)Katie Shpanskaya(Stanford University)Francis G. Blankenberg(Stanford University)Jayne Seekins(Stanford University)Timothy J. Amrhein(Duke University)David A. Mong(University of Colorado Denver)Safwan S. Halabi(Stanford University)Evan J. Zucker(Stanford University)Andrew Y. Ng(Stanford University)Matthew P. Lungren(Stanford University)

View via DOI Full-text PDF

Abstract

In this study, we developed and validated a deep learning algorithm that classified clinically important abnormalities in chest radiographs at a performance level comparable to practicing radiologists. Once tested prospectively in clinical settings, the algorithm could have the potential to expand patient access to chest radiograph diagnostics.

COVID-19 diagnosis using AI Lung Cancer Diagnosis and Treatment Radiology practices and educationMedicineChest radiographRadiologyReceiver operating characteristicConvolutional neural networkRadiographyArtificial intelligencePleural effusionDeep learningMedical diagnosis

MeSH terms

RadiologistsDeep LearningClinical CompetenceDiagnosis, Computer-AssistedHumansPneumoniaPredictive Value of TestsRadiographic Image Interpretation, Computer-AssistedRetrospective StudiesRadiography, ThoracicReproducibility of Results