A Cross-Sectional Technology Acceptance Study of an AI CAD System in a Breast Screening Unit

Sammanfattning: In January 2021, one of the first large-scale implementations of an artificial intelligent computer-aided detection system (AI CAD) for detecting breast cancer was implemented at Capio S:t Göran hospital in Stockholm. AI CAD for detecting breast cancer is promising, however, it can only be a successful implementation if it is accepted by the end-users. This study examines and evaluates factors critical to the acceptance of the AI CAD, prior to the radiology professionals using it by applying the third version of the Technology Acceptance Model, that is TAM3. A questionnaire was designed and distributed accordingly to 28 professionals at the hospital’s breast screening unit. The quantitative data collected were further analyzed using the statistical tool SPSS. The empirical findings concluded that the intention to use the AI CAD was influenced directly by the perceived usefulness and indirectly by image, job relevance, and perceived ease of use. In addition, the association between subjective norm and image was shown to be significant. This study further revealed two new associations, contrary to what TAM3 postulates, the first one being between image and behavioral intention and the second one being between job relevance and behavioral intention. Organizational support, system-related activities, and information and communication are interventions suggested based on the findings in this study, in which the breast screening unit at S:t Göran should tap into to enhance the acceptance of the AI CAD system.