Receive a weekly summary and discussion of the top papers of the week by leading researchers in the field.

In Academic radiology ; h5-index 0.0

RATIONALE AND OBJECTIVES : Artificial intelligence (AI) is playing a growing role in the field of radiology. This article seeks to help readers quantify its impact when put into practice, using a lung nodule flagger as an example.

MATERIALS AND METHODS : The one-time and ongoing costs associated with AI are explored. Costs are divided into three categories: direct costs, costs associated with operational changes, and downstream costs. Examples of each are provided.

RESULTS : A framework for estimating the financial impact of AI is provided.

CONCLUSION : The impact of AI is quantifiable, but estimates of its financial impact may not be portable across contexts. Different organizations may implement AI in different ways due to differences in clinical practices. Furthermore, different organizations have different hurdle rates for their investments. Finally, international cost-effectiveness analyses may not be generalizable due to differences in both practice patterns and the valuation placed upon quality. When quantifying the impact of AI, organizations should consider relying upon pilots and data from other similarly-situated organizations.

Powell Adam C

2020-Jan

AI, Artificial intelligence, Cost, Financial impact, Radiology, Time-value of money