AI Improves Efficiency of FDG-PET/CT Interpretations of Lymphoma

Background: Fluorine-18 fluorodeoxyglucose (FDG)-PET/CT imaging is considered the gold standard for the image evaluation of lymphoma. In light of the increasingly limited radiology workforce, providing efficient and accurate interpretation of these studies by artificial intelligence (AI) methods may be helpful. Objective: To determine whether use of an AI-assisted PET/CT reporting workflow could improve the interpretation of pretreatment staging FDG-PET/CT examinations in high-grade lymphoma. Design: Retrospective study. Methods: The authors used a crossover design using an AI-assisted PET/CT reading program compared to the routine clinical reporting workflow process. FDG-PET/CT scans of patients with high-grade lymphoma cases were obtained. An AI-assisted PET/CT reading program was compared to the routine clinical reporting workflow process. The FDG-PET/CT scans were performed with 6 hours of fasting before FDG administration. Low-dose CT was obtained for attenuation correction. Nin more...