AI Performance for Nodule Volume Doubling Time in the follow-up of the UKLS Lung Cancer Screening Study Compared to Expert Consensus and Histological Validation

This study evaluated the performance of an AI system in measuring nodule volume doubling time (VDT) during follow-up in the UK Lung Cancer Screening (UKLS) study. A total of 154 nodules from 133 participants were analyzed, including 47 malignant nodules. Aview LCS (Coreline Soft) was compared with expert consensus and histological validation. The AI demonstrated high agreement with expert assessment (ICC 0.94) and achieved 89.4% sensitivity and 94.4% specificity for identifying nodules with VDT under 400 days. It also showed strong performance in malignancy prediction with an AUC of 0.92. These results indicate that AI-based VDT measurement is a reliable approach for nodule management and follow-up in lung cancer screening programs.