PP01.58 Multi City Opportunistic Screening of Lung Nodules amidst COVID-19

Early diagnosis of lung cancer can substantially reduce cancer related deaths. Missed nodules have been reported in literature, as is the benefits of incidental nodule detection and management. The use of AI in flagging incidental nodules was explored on CT scans taken for COVID and this has been submitted to CHEST, and RSNA during various stages of the processing with respect to sites. 2502 scans taken during the second and third wave of COVID at two specialist radiology chain (5 sites) were processed by qCT Lung, an AI capable of detecting and characterizing nodules. Radiologist report of the cases flagged by qCT were searched for findings suggestive of cancer. Cases for which nodule was not reported were re-read by an independent radiologist with AI assistance on online portal. They were asked to either confirm or reject the flag, rate the nodule for malignancy potential if confirmed or provide alternate finding if rejected (See Figure). 1737 nodules were flagged in 673 (26.9%). 80 were noted in radiology report. Of the 593 scan re-read by radiologist, 115 (19.4%) was confirmed as having nodules. 233 nodules were found in these scans. The median size (longest diameter) of these nodules was 15 mm (range: 6 – 48) and 64.4% were solid. The most common cause of rejection was ground glass opacities. 94 of 115 scans had a rating of 1 or 2. 19 were called unsure and two was probably malignant. Two cases were called cancer by both AI and radiology report. Missed cancer is medico-legal problem. In this study it cannot be assumed that the radiologist missed 115 nodules. These radiologists might have spotted the nodule but not reported it because of their perceived risk and this is supported by second reader. However, 21 cases would have warranted follow-up.