Machine learning in cardiovascular radiology: ESCR position statement on design requirements, quality assessment, current applications, opportunities, and challenges

Machine learning offers great opportunities to streamline and improve clinical care from the perspective of cardiac imagers, patients, and the industry and is a very active scientific research field. In light of these advances, the European Society of Cardiovascular Radiology (ESCR), a non-profit medical society dedicated to advancing cardiovascular radiology, has assembled a position statement regarding the use of machine learning (ML) in cardiovascular imaging. The purpose of this statement is to provide guidance on requirements for successful development and implementation of ML applications in cardiovascular imaging. In particular, recommendations on how to adequately design ML studies and how to report and interpret their results are provided. Finally, we identify opportunities and challenges ahead. While the focus of this position statement is ML development in cardiovascular imaging, most considerations are relevant to ML in radiology in general. Key Points • Development and clinical implementation of machine learning in cardiovascular imaging is a multidisciplinary pursuit. • Based on existing study quality standard frameworks such as SPIRIT and STARD, we propose a list of quality criteria for ML studies in radiology. • The cardiovascular imaging research community should strive for the compilation of multicenter datasets for the development, evaluation, and benchmarking of ML algorithms.

Cardiac imaging procedures and the COVID-19 pandemic: recommendations of the European Society of Cardiovascular Radiology (ESCR)

The severe acute respiratory syndrome coronavirus 2019 (SARS-CoV-2) pandemic currently constitutes a significant burden on worldwide health care systems, with important implications on many levels, including radiology departments. Given the established fundamental role of cardiovascular imaging in modern healthcare, and the specific value of cardiopulmonary radiology in COVID-19 patients, departmental organisation and imaging programs need to be restructured during the pandemic in order to provide access to modern cardiovascular services to both infected and non-infected patients while ensuring safety for healthcare professionals. The uninterrupted availability of cardiovascular radiology services remains, particularly during the current pandemic outbreak, crucial for the initial evaluation and further follow-up of patients with suspected or known cardiovascular diseases in order to avoid unnecessary complications. Suspected or established COVID-19 patients may also have concomitant cardiovascular symptoms and require further imaging investigations. This statement by the European Society of Cardiovascular Radiology (ESCR) provides information on measures for safety of healthcare professionals and recommendations for cardiovascular imaging during the pandemic in both non-infected and COVID-19 patients.

Detection of coronary artery calcification with nontriggered computed tomography of the chest

Objective: To evaluate the accuracy of visual analysis and of the coronary artery calcium (CAC) score in nontriggered computed tomography (CT), in comparison with that of the CAC score in electrocardiogram-triggered CT, in identifying coronary calcification. Materials and Methods: A total of 174 patients for whom CT was indicated for CAC scoring underwent nontriggered and triggered CT in a 64-channel multislice scanner, in a single session without a change in position. The images were interpreted by a radiologist with seven years of experience in thoracic and cardiovascular radiology. The measurement of coronary calcium was carried out by three methods: CAC score with dedicated software in nontriggered CT, CAC score with dedicated software in triggered CT, and visual analysis without dedicated software in nontriggered CT. Results: In nontriggered CT, the CAC score presented an accuracy of 95.98% (95% CI: 91.93-98.04). The visual analysis showed an accuracy of 97.13% (95% CI: 93.45-98.77). Conclusion: Nontriggered CT showed excellent accuracy in the identification and exclusion of coronary calcification, either the CAC score was determined with dedicated software or through visual analysis.