Real-world evidence (RWE) is a term applied to describe the use of data collected from routine clinical practice, as opposed to data collected in clinical trials, which prioritized environmental control to understand the potential effects of an intervention or exposition on variables dependent on results . Essentially, RWE covers the use of data from medical records, data collected through specific tests validated for clinical practice, and routine exams, among other sources . In recent years, RWE has gained widespread notoriety in the field of medical imaging for the study of the safety and efficacy of imaging modalities and techniques in real-world environments, above all, there is a significant interest in the use of RWE to understand the performance of imaging modalities and techniques into clinical practice environment . These data can result in research that contributes to the development of a strategy for clinical decision-making and supports the development and implementation of new supporting technologies .One of the adjuvant technologies for the use of RWE is artificial intelligence (AI) in medical imaging . Traditionally, the use of AI models has been widely used as an auxiliary tool for radiologists to interpret medical images. The use of RWE can be used to evaluate the performance of these algorithms in a real-world environment, respectively contributing to increase accuracy, workflow, and resources used in the diagnosis of AI algorithms . However, RWE is not without limitations, such as AI models, proper data recording, standardization, data security and privacy, and generalization [4-5].
I. Data recording: RWE data are subject to the influence of biases related to the instrument (i.e., protocol reproducibility and validation), evaluator (i.e., level of experience with the instrument), environmental factors (number of patients in daily follow-up, hospital flow, location of hospital).
II. Privacy and security: RWE data contains sensitive personal information; therefore, continuous strategies have been developed to ensure the anonymization of this data.
III. Data analysis: the volume of data generated in RWE studies, especially when applied to the medical imaging scenario, encompasses a considerable challenge regarding data processing, and database creation, resulting in an increase in human resources for analysis and search for data-related insights.
IV. Ethics: data from RWE may raise issues related to informed consent, sharing, and control of data, as well as potential conflicts of interest.
Artificial intelligence (AI) is improving RWE in medical imaging in a number of ways. AI algorithms can be trained to automatically extract and analyze data from medical images, reducing the need for manual interpretation and increasing the efficiency of data collection. Additionally, AI can be used to identify patterns and relationships in the data that may not be apparent to the human eye, leading to new insights and discoveries. AI can also be used to develop new imaging modalities and to improve image quality, making the data more reliable and informative. These developments are helping to make RWE more widely available and more useful for improving patient care.
In general, the use of RWE data has gained notoriety within the medical imaging and AI scenario, with the potential to provide valuable information about the real performance of AI algorithms, contributing to the development of technologies that can facilitate diagnosis, reduce workload, corroborate the increase in survival and quality of life of patients and health professionals involved in attention and care. However, despite the extensive development of RWE, there are still considerable gaps in the generalization and privacy of these data. In this way, efforts have been made to fill these gaps [1-5].
 Gauthier MP, Law JH, Le LW, Li JJN, Zahir S et al. Automating Access to Real-World Evidence. JTO Clin Res Rep. 2022 May 17;3(6):100340.
 Lin D, Xiong J, Liu C, Zhao L, Li Z, et al. Application of Comprehensive Artificial intelligence Retinal Expert (CARE) system: a national real-world evidence study. Lancet Digit Health. 2021 Aug;3(8):e486-e495.
 Wellnhofer E. Real-World and Regulatory Perspectives of Artificial Intelligence in Cardiovascular Imaging. Front Cardiovasc Med. 2022 Jul 22;9:890809.
 Mudgal SK, Agarwal R, Chaturvedi J, Gaur R, Ranjan N. Real-world application, challenges and implication of artificial intelligence in healthcare: an essay. Pan Afr Med J. 2022 Sep 2;43:3.
 Passamonti F, Corrao G, Castellani G, Mora B, Maggioni G, et al.. The future of research in hematology: Integration of conventional studies with real-world data and artificial intelligence. Blood Rev. 2022 Jul;54:100914.