Artificial intelligence (AI) could probably be used as a second reader in lung cancer screening just aswith mammography, Chiara Romei from Pisa University Hospital, Italy, told Ben Giese at ECR2024 last February in Vienna.
‘Lung cancer screening demonstrates excellent results and reduced mortality. The problem now is to figure out how we use it in clinical practice in patient pathway,’ she said. ‘The possibility to have a second reader integrated in PACS that uses automatic detection of lung nodules based on AI could be very interesting.’
AI could also be used as a detection tool to start training young radiologists and residents to increase confidence, she added. ‘For example for chest x-ray, they can use the tools that use detection of pneumothorax or plural effusion or consolidation to increase their confidence. Or for chest CT, they can use the tool to increase their confidence and be trained,’ she said.
One of the drawbacks of using AI as a second reader in both x-ray and CT could be to rely too much on the software. A problem could be ‘that the radiologists do what the computer tells them to do, without using human intelligence (in their) decision making.’