Can we revolutionize diagnostic imaging by keeping Pandora's box closed?

Area of Science:

• Medical Imaging
• Artificial Intelligence
• Healthcare Management

Background:

• Incidental imaging findings present a significant challenge in healthcare, frequently leading to low-value and potentially harmful patient care.
• Current practices struggle with managing incidental findings once detected, as they often cannot be ignored by healthcare professionals.

Purpose of the Study:

• To reflect on current practices for managing incidental imaging findings.
• To propose and discuss a novel strategy for preemptively addressing incidental findings in diagnostic imaging.

Main Methods:

• The article is an opinion piece, reflecting on existing practices.
• It proposes a new strategy centered on preventing the visualization of incidental findings.
• Deep learning algorithms are suggested as a method to achieve this prevention.

Main Results:

• The proposed strategy aims to 'keep Pandora's box closed' by not visualizing incidental findings.
• This approach leverages deep learning algorithms for pre-emptive management.
• The concept has the potential for significant impact on diagnostic imaging practices.

Conclusions:

• Incidental findings pose a considerable problem, often resulting in low-value care.
• A novel strategy using deep learning to avoid visualizing incidental findings is proposed.
• This approach could profoundly change diagnostic imaging by preventing unnecessary interventions.