Evaluation of the impact of novel digital PET/computed tomography reconstructions in Herder score calculations for the assessment of solitary pulmonary nodules on 2-[18F]fluoro-2-deoxy-D-glucose PET/computed tomography

PURPOSE: The 4-point visual 2-[ 18 F]fluoro-2-deoxy-D-glucose ([ 18 F]FDG) PET/computed tomography (CT) 'Herder' score, in combination with clinical factors, is used to assess malignancy risk of solitary pulmonary nodules (SPNs) and guide management. We aim to assess the impact of four digital reconstructions on Herder visual score, Herder model probability, and subsequent management of SPNs. METHODS: [ 18 F]FDG PET/CT examinations for SPN assessment, performed on a Siemens Biograph Vision PET/CT, were retrospectively included (398 reconstructions; Gauss 6.0, Gauss 4.5, Gauss 4.5 with data-driven gating, and All Pass). Eight expert readers assessed anonymised reconstructions randomly (four pairs, 25 cases/100 reconstructions per pair). The Herder visual score was documented, and the Herder model probability score for each reconstruction was calculated using clinical parameters. A single observer performed semiquantitative analysis (SUV max and target-to-background ratios). RESULTS: 73/100 cases showed Herder visual score concordance across all four reconstructions. Although the visual score changed for at least one reconstruction in 27 cases, the risk stratification based on the Herder model probability remained identical in 15/27. In 12 cases, changes in Herder visual score altered risk stratification. However, this did not change the subsequent management plan for any patient. Semiquantitative outputs were significantly different between Gauss 6.0 and All Pass ( P < 0.001) and Gauss 4.5 and All Pass ( P = 0.03-0.05) with no significant difference between the other reconstructions. CONCLUSION: There is excellent concordance in Herder based SPN risk stratification across all four digital PET reconstructions. Although different reconstructions can alter risk stratification, this did not impact management in our cohort.