An Introductory Guide to Artificial Intelligence in Interventional Radiology: Part 2: Implementation Considerations...

Area of Science:

• Medical Imaging and Interventional Radiology
• Artificial Intelligence in Healthcare

Background:

• Artificial intelligence (AI) integration into interventional radiology (IR) is rapidly advancing.
• AI applications in IR encompass software as a medical device and AI-integrated hardware.
• The adoption of AI necessitates careful consideration of potential risks and benefits for patients and clinicians.

Purpose of the Study:

• To outline the challenges and opportunities presented by AI in interventional radiology.
• To describe a framework for evaluating AI medical devices based on risk.
• To provide practical guidance for the clinical deployment of AI in IR.

Main Methods:

• A review of current AI applications in IR.
• Development of a risk-based hierarchy for AI implementation.
• Creation of a deployment checklist for clinical IR environments.

Main Results:

• AI in IR offers significant opportunities but requires careful management of risks.
• A hierarchy of risk is proposed to guide the evaluation of AI medical devices.
• A practical checklist is provided to facilitate safe AI deployment in clinical settings.

Conclusions:

• The integration of AI in IR demands a proactive and risk-aware approach.
• Continuous evolution of AI necessitates adaptive regulatory and evaluation strategies.
• Ensuring patient safety is paramount throughout the AI lifecycle in IR.