Avoid common mistakes on your manuscript.
Dear Reader,
For this July 2023 issue, I have selected the paper by Dr Miguel Nobre Menezes and co-authors from the Cardiovascular Center of the University of Lisbon in Portugal [1]. The reasons for me to select this paper are: (1) my personal long-term interest in QCA in general; and (2) the progress that is being made also in this field using Artificial Intelligence (AI), which is developing at an amazing pace in all kinds of applications.
In this paper the authors validated an earlier developed AI approach using 123 regions of interest from 117 images across 90 patients selected on retrospective data from four different centers; an established QCA system was used as reference. For each selected lesion, a single end-diastolic frame with clear outline definition of the vessel and target lesion was selected. More than one segment per patient and/or image could be used. Images with poor image quality were excluded from the analysis. Fig. 1 below illustrates the overview of the segmentation and analysis process.
Overall, the results were found to be very good with overlap accuracy between the original and segmented images of 99.9%, a sensitivity of 95.1% and a Dice score of 94.8%. The authors concluded that the AI model was capable of accurate CAG segmentation across multiple performance metrics.
There are limitations to these results of course, as also mentioned in the paper by the authors. As next steps, further improving and expanding the model is an essential requirement, whereby large, annotated data sets are needed. The proof of the pudding in the end is of course, when such systems are used worldwide in daily research and clinical practices, and that will still require significant efforts. Not to forget also, the regulatory requirements and acceptance by the field of fully automated systems, which are not to be under-estimated. However, this paper should be seen as an example of where the field is moving to.
Anyway, with this July 2023 choice, I would like to thank you, our readership, again for your interest and would like to wish you much reading pleasure with this paper and all the other very interesting papers published in this July 2023 issue of the International Journal of Cardiovascular Imaging.
Johan HC Reiber, PhD.
Editor-in-chief.
e-mail: J.H.C.Reiber@lumc.nl.
References
Nobre Menezes M, Silva JL, Silva B et al (2023) Coronary X-ray angiography segmentation using Artificial Intelligence: a multicentric validation study of a deep learning model. Int J Cardiovasc Imaging. https://doi.org/10.1007/s10554-023-02839-5
Author information
Authors and Affiliations
Corresponding author
Additional information
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
About this article
Cite this article
Reiber, J.H. Coronary X-ray angiography segmentation using Artificial Intelligence: a multicentric validation study of a deep learning model. Int J Cardiovasc Imaging 39, 1217–1219 (2023). https://doi.org/10.1007/s10554-023-02889-9
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10554-023-02889-9