Comparing a Top-Down and a Bottom-Up Approach for Implementing AI in Radiology Practice

Area of Science:

• Health Services Research
• Health Informatics
• Artificial Intelligence in Healthcare

Background:

• Healthcare organizations are increasingly exploring the integration of commercial artificial intelligence (AI) algorithms.
• The implementation of AI in healthcare presents unique challenges and requires strategic planning.
• Understanding different implementation pathways is crucial for successful technology adoption.

Purpose of the Study:

• To examine and compare two distinct strategies for implementing commercial AI algorithms in Norwegian health regions.
• To identify the key characteristics of top-down, research-driven, and bottom-up, innovation-focused implementation approaches.
• To assess the knowledge generated by each strategy for broader applicability in healthcare settings.

Main Methods:

• Comparative case study analysis of two Norwegian health regions.
• Qualitative examination of AI algorithm implementation strategies.
• Documentation of regional approaches, decision-making processes, and outcomes.

Main Results:

• One health region adopted a top-down, research-driven strategy for AI implementation.
• The second health region utilized a bottom-up, innovation-focused strategy for AI implementation.
• Both strategies generated distinct but valuable knowledge regarding AI implementation in healthcare.

Conclusions:

• There is no single optimal approach for implementing commercial AI algorithms in healthcare.
• Both top-down and bottom-up strategies can effectively facilitate AI adoption.
• The knowledge gained from diverse implementation strategies is transferable and beneficial for other healthcare settings.