Multiple Use Case Analysis of an AI Hospital Using the AIDAF and Intelligence Amplification Design Canvas

In March 2023, the five-year AI Hospital Program in Japan was completed. Various AI-applications have been implemented in four model AI Hospitals. This paper contains the results of a multiple use case analysis of an AI Hospital using the AIDAF and the intelligence amplification design canvas. The aim of this research paper is two-fold to: 1) evaluate the utility of the AIDAF and intelligence amplification design canvas to analyse implemented AI-applications, and 2) conduct a multiple use case analysis in an AI Hospital. A focus group was created and developed a three-step process to analyse nine use cases and related AI-applications that have been implemented at Keio University Hospital using the AIDAF and intelligence amplification design canvas. First, each use case was analysed using the four guiding principles of the intelligence amplification design canvas. Second, conceptual representations were created for each use case. Third, an aggregated view of the AI Hospital was created. The results and findings were discussed in five focus group meetings. This study provides empirical support that the AIDAF and the intelligence amplification design canvas can effectively support the individual analysis of implemented AI-applications as well as conducting a multiple use case analysis. As such, this study provided insight in the current state of the AI Hospital and identified opportunities for extensions and future development in the second term of the AI Hospital Program and Open Healthcare Platform 2030 consortium. The current study was mainly based on qualitative analysis and comparison implemented AI-applications within a single organization and did not include in-depth user research. Future research may contribute to the analysis of similar use cases in other (AI-)hospitals to conduct a cross use case analysis.