Will robots take our jobs? A qualitative study about technology acceptance of artificial intelligence among radiologists

Abstract: Information technologies coming from non-healthcare sectors have proven to be difficult to translate, implement and adopt in the healthcare sector. One important factor that influence the adoption process is the attitude towards the technology, what is called the acceptance of technology. In healthcare, the field of radiology will most likely be transformed by artificial intelligence (AI), due to the fact that diagnostic imaging is one of the most apparent applications for AI. Therefore, it is interesting to understand acceptance of AI in radiology. Even though there are a lot of articles discussing the potential of AI, there seems to be a lack of papers covering the acceptance of AI in radiology. With this in mind, the purpose of this study is to understand what factors affect the acceptance of AI among radiologists. For this purpose, 18 semi-structured interviews with physicians (radiologists and neurologists) and decision makers within radiology were conducted. An inductive thematic analysis of the empirical data was performed and then assessed for its fit to the Technology Acceptance Model 2. The authors find a relatively good fit to the existing framework. Furthermore, the study indicates how concerns of job relevance and output quality, as well as a new determinant called "uncertainty", has a major impact on the radiologists' acceptance of AI. Specifically, the study's empirically found antecedents; control and applicability, are experienced as important. In addition, the findings support that subjective norm and image may affect acceptance of AI to a greater extent than they have affected acceptance of previous technologies.