AI integrations with lung cancer screening: Considerations in developing AI in a public health setting

Area of Science:

• Radiology and Medical Imaging
• Public Health
• Artificial Intelligence in Medicine

Background:

• Lung cancer screening (LCS) in older, tobacco-exposed populations reveals incidental findings like emphysema and coronary artery disease (CAD).
• Established protocols exist for interventions based on these findings, particularly coronary calcium.
• The increasing volume of LCS necessitates a review of screening infrastructure development.

Purpose of the Study:

• To review issues in developing effective chest CT screening infrastructure.
• To address the integration of advanced imaging analysis, including artificial intelligence (AI).
• To balance the needs of AI tool developers with public health requirements for equitable and accessible screening.

Main Methods:

• Review of current LCS practices and emerging diagnostic innovations.
• Analysis of imaging methodology requirements for reliable quantitative characterization of lung nodules, emphysema, and coronary artery disease.
• Consideration of AI's role in chest CT screening and its implications for precision, reproducibility, and cost.

Main Results:

• LCS success relies on precise characterization of lung nodules (3-15 mm).
• Similar imaging quality is essential for reliable quantification of emphysema and coronary artery disease.
• AI is significantly impacting medical imaging, including chest CT screening.

Conclusions:

• Developing robust chest CT screening infrastructure is timely due to technological advancements and increasing participant numbers.
• AI integration in chest CT screening requires careful consideration to ensure precision, reproducibility, and cost-effectiveness.
• The Alliance for Global Implementation of Lung and Cardiac Early Disease Detection and Treatment (AGILE^DxRx) aims to facilitate equitable, high-quality, multi-disciplinary chest CT screening using advanced computational tools.