Hopkins study suggests that computer programs can help radiologists diagnose pancreatic cancer.

Researchers from the Sol Goldman Pancreatic Cancer Research Center at Johns Hopkins recently participated in a study that applied the new computer tool called “radiomics” to CT scans of the pancreas. The study, reported in the April issue of the American Journal of Roentgenology (AJR) showed that radiomics features can be used to identify pancreatic cancer in CT scans. The Hopkins team, which included radiologists, pathologists, computer scientists, and cancer biologists, analyzed computed tomography (CT) scans from 380 patients – 190 with pancreatic cancer and 190 healthy controls. Their initial analysis revealed 40 radiomics features which were then used to classify a set of validation scans as either cancer or normal. This new classifier had >99% accuracy and correctly identified all cases of pancreatic cancer, suggesting it could be a powerful complement to human review of imaging data. The study suggests that computer programs can be developed that will help radiologists diagnose pancreatic cancer more accurately. To learn more, visit: https://www.ajronline.org/doi/full/10.2214/AJR.18.20901

Comments are closed.