Scaling AI Projects for Radiology - Causes and Consequences.

Artificial intelligence (AI) for radiology has the potential to handle an ever-increasing volume of imaging examinations. However, the implementation of AI for clinical practice has not lived up to expectations. We suggest that a key problem with AI projects in radiology is that high expectations associated with new and unproven AI technology tend to scale the projects in ways that challenge their anchoring in local practice and their initial purpose of serving local needs. Empirically, we focus on the procurement of an AI solution for radiology practice at a large health trust in Norway where it was intended that AI technology would be used to process the screening of images more effectively. Theoretically, we draw on the information infrastructure literature, which is concerned with scaling innovative technologies from local settings, with a limited number of users, to broad-use contexts with many users.

Correction: Elaborating Models of eHealth Governance: Qualitative Systematic Review.

[This corrects the article DOI: 10.2196/17214.].

Elaborating Models of eHealth Governance: Qualitative Systematic Review.

BACKGROUND: Large-scale national eHealth policy programs have gained attention not only for benefits but also for several unintended consequences and failed expectations. Given the complex and mixed accounts of the results, questions have been raised on how large-scale digitalization programs are governed to reach health policy goals of quality improvement and equal access along with necessary digital transformations. In this qualitative systematic review, we investigate the following question: How is governance implemented and considered in the studies included in the qualitative review? OBJECTIVE: The aim of this study is to arrive at informed and recognizable conceptualizations and considerations of models of governance connected to eHealth, as presented and discussed in the scientific literature. In turn, we hope our results will help inform the discussion of how to govern such processes to obtain collectively negotiated objectives. METHODS: A qualitative systematic review is a method for integrating or comparing with the findings from qualitative studies. It looks for "themes" or "constructs" that lie in or across individual qualitative studies. This type of review produces a narrative synthesis with thematic analysis and includes interpretive conceptual models. The goal is an interpretation and broadens the understanding of a particular phenomenon. We searched the PubMed database using predefined search terms and selected papers published from 2010 onwards. We specified the criteria for selection and quality assessment. RESULTS: The search returned 220 papers. We selected 44 abstracts for full-text reading, and 11 papers were included for full-text synthesis. On the basis of the 11 papers, we constructed four governance models to categorize and conceptualize the findings. The models are political governance, normally depicting top-down processes; medical governance, which normally depicts bottom-up processes; the internet and global model, emphasizing international business strategies coupled with the internet; self-governance, which builds upon the development of the internet and Internet of Things, which has paved the way for personal governance and communication of one's own health data. CONCLUSIONS: Collective negotiations between the nation-state and global policy actors, medical and self-governance actors, and global business and industry actors are essential. Technological affordances represent both positive and negative opportunities concerning the realization of health policy goals, and future studies should scrutinize this dynamic.