BackgroundEstimation of the risk of malignancy in pulmonary nodules detected by CT is central in clinical management. The use of artificial intelligence (AI) offers an opportunity to improve risk prediction. Here we compare the performance of an AI algorithm, the lung cancer prediction convolutional neural network (LCP-CNN), with that of the Brock University model, recommended in UK guidelines.MethodsA dataset of incidentally detected pulmonary nodules measuring 5–15 mm was collected retrospectively from three UK hospitals for use in a validation study. Ground truth diagnosis for each nodule was based on histology (required for any cancer), resolution, stability or (for pulmonary lymph nodes only) expert opinion. There were 1397 nodules in 1187 patients, of which 234 nodules in 229 (19.3%) patients were cancer. Model discrimination and performance statistics at predefined score thresholds were compared between the Brock model and the LCP-CNN.ResultsThe area under the curve for LCP-CNN was 89.6% (95% CI 87.6 to 91.5), compared with 86.8% (95% CI 84.3 to 89.1) for the Brock model (p≤0.005). Using the LCP-CNN, we found that 24.5% of nodules scored below the lowest cancer nodule score, compared with 10.9% using the Brock score. Using the predefined thresholds, we found that the LCP-CNN gave one false negative (0.4% of cancers), whereas the Brock model gave six (2.5%), while specificity statistics were similar between the two models.ConclusionThe LCP-CNN score has better discrimination and allows a larger proportion of benign nodules to be identified without missing cancers than the Brock model. This has the potential to substantially reduce the proportion of surveillance CT scans required and thus save significant resources.

Objectives: The use of cross-sectional imaging in clinical medicine has been a major step forward in the management of many conditions but with that comes the increasing demand on resources and the detection of other potentially significant findings. This, in the context of a shortage of skilled radiologists, means that new ways of working are important. In thoracic CT, pulmonary nodules are a significant challenge because they are so common. Poor and inconsistent management can both cause harm to patients and waste resources so it is important that the latest guidelines are followed. The latter mandate the use of semi-automated volumetry that allows more precise management but is time-consuming. Methods: Reporting radiographers were iteratively trained in the use of semi-automated volumetry for pulmonary nodules by experienced thoracic radiologists. Once trained in this specific aspect, radiographers completed reporting of pulmonary nodules, checked by radiologists. Results: Radiographer reporting reduced radiologist time in reporting nodules and measuring their volume. Most of the volumetry was completed prior to the multidisciplinary meeting. This facilitated an increase in the number of patients discussed in 60 min from 15 to 22. Radiographers failed to detect few nodules, although a second read by radiologists is required in any case for other aspects of the reporting. Conclusion: Reporting radiographers, working with radiologists in a supportive setting, can deliver the radiology in a lung nodule pathway, reducing the time commitment from radiologists and the pulmonary nodule multidisciplinary team members, whilst using this as an opportunity to conduct research.

No abstract

No abstract