Deep Learning Detection of Direct and Indirect Imaging Findings Associated with Pancreatic Cancer at Contrast-enhanced and Noncontrast CT

The authors aimed to develop and evaluate deep learning (DL) models for detecting both direct and indirect imaging findings associated with pancreatic cancer (PC) on noncontrast and contrast-enhanced CT images. The study found that these DL models demonstrated diagnostic performance comparable to or better than that of experienced physicians, particularly in identifying smaller pancreatic cancers. Overall, the models effectively detected key imaging findings and diagnosed PC, suggesting their potential utility in clinical practice.