To systematically evaluate CE-marked AI products for lung nodule analysis, characterizing their capabilities and assessing their coverage of nodule management tasks, including specific evaluation criteria.
Key Findings:
AI products vary in their support for nodule management tasks, with some fully supporting while others only partially or not at all, impacting clinical decision-making.
Task coverage scores were calculated to assess the functional overlap of AI products with established nodule management recommendations, highlighting areas for improvement.
Interpretation:
The study highlights the potential of AI to assist in lung cancer screening but also reveals critical gaps in product capabilities that need to be addressed for effective clinical integration.
Limitations:
The analysis was limited to CE-marked products available as of January 1, 2025, potentially excluding newer solutions that may offer improved capabilities.
Vendor responses were not guaranteed, leading to reliance on publicly available data for some assessments, which may affect the reliability of findings.
Conclusion:
The evaluation underscores the need for systematic assessment of AI tools in lung cancer screening to ensure they meet clinical requirements and support effective nodule management, addressing identified gaps for future research.
