Deep Learning for Detection of Pneumothorax and Pleural Effusion on Chest Radiographs: Validation Against Computed Tomography, Impact on Resident Reading Time, and Interreader Concordance

• Published in: Journal of thoracic imaging Vol. 39; no. 3; p. 185
• Main Authors: Tejani, Ali, Dowling, Thomas, Sanampudi, Sreeja, Yazdani, Rana, Canan, Arzu, Malja, Elona, Xi, Yin, Abbara, Suhny, Peshock, Ron M, Kay, Fernando U
• Format: Journal Article
• Language: English
• Published: United States 01.05.2024
• Subjects: Artificial Intelligence; Deep Learning; Humans; Pleural Effusion - diagnostic imaging; Pneumothorax - diagnostic imaging; Radiography, Thoracic; Retrospective Studies; Tomography, X-Ray Computed

To study the performance of artificial intelligence (AI) for detecting pleural pathology on chest radiographs (CXRs) using computed tomography as ground truth. Retrospective study of subjects undergoing CXR in various clinical settings. Computed tomography obtained within 24 hours of the CXR was use...