Dimensions of Human-Machine Combination: Prompting the Development of Deployable Intelligent Decision Systems for...

Area of Science:

• Healthcare technology
• Human-computer interaction
• Artificial intelligence

Background:

• A significant gap exists between the conception and implementation of Artificial Intelligence (AI) in clinical settings.
• Few prospective situated studies focus on the practical application of AI in healthcare.

Purpose of the Study:

• To advocate for increased use of situated design and evaluation to bridge the AI implementation gap in healthcare.
• To introduce a novel framework, contextual dimensions of combination, for analyzing human-machine decision-making processes.
• To identify key dimensions for designing user-centered AI in clinical contexts.

Main Methods:

• Literature review to identify the scarcity of prospective situated studies in clinical AI.
• Development of a conceptual framework: contextual dimensions of combination.
• Identification of eight initial dimensions within this framework: participating agents, control relations, task overlap, temporal patterning, informational proximity, informational overlap, input influence, and output representation coverage.

Main Results:

• The literature shows a lack of prospective situated studies for clinical AI.
• A novel framework, contextual dimensions of combination, is proposed to structure the analysis of human-machine interaction.
• Eight key dimensions are identified to guide the design and evaluation of AI systems in healthcare.

Conclusions:

• Situated design and evaluation are crucial for successful clinical AI implementation.
• The contextual dimensions of combination framework provides a structured approach to address challenges in human-AI collaboration.
• Awareness of these dimensions will drive the development of more effective and user-centered AI designs, ultimately benefiting patients and society.