Convolutional neural network for automatic detection and characterization of abdominal aortic aneurysm.

Abstract

We read with great interest the recent publication of Camara et al. on the development of an artificial intelligence (AI)-derived method to enable an automatic detection of abdominal aortic aneurysm (AAA) (1Camara JR TR, Pop A, Matthew BS, Shedd P, Brandon S, Dobrowski BS, Cole J, Knox BS, Abou-Zamzam A, Sharon C, Kiang Journal of Vascular Surgery Cases and Innovative techniquesGoogle Scholar). Using datasets composed of 200 computed tomography angiography (CTA) from patients with AAA and 200 matched control patients with non-aneurysmal infrarenal aorta, the authors developed a method based on convolutional neural network (CNN) and testing demonstrated a robust accuracy of the model (99.1% with an area under the curve of 0.99). Their results point to the interest of such application for the screening of AAA. The VGG-16 neural network architecture was used to develop the AAA detection system and transfer learning was applied to the neural network. It would be interesting to know what was the accuracy of the CNN before transfer learning to show what was the added value of the transfer learning in the pipeline. The use of CNN to classify aneurysmal from non-pathological aortas has been so far poorly reported (2Mohammadi S. Mohammadi M. Dehlaghi V. Ahmadi A. Automatic Segmentation, Detection, and Diagnosis of Abdominal Aortic Aneurysm (AAA) Using Convolutional Neural Networks and Hough Circles Algorithm.Cardiovasc Eng Technol. 2019; 10: 490-499Crossref PubMed Scopus (27) Google Scholar). Nevertheless, it is to note that several studies have recently demonstrated the interest of CNN to develop a fully automatic segmentation of AAA (3Lareyre F. Adam C. Carrier M. Raffort J. Automated Segmentation of the Human Abdominal Vascular System Using a Hybrid Approach Combining Expert System and Supervised Deep Learning.J Clin Med. 2021; 10Crossref Scopus (9) Google Scholar, 4Caradu C. Spampinato B. Vrancianu A.M. Berard X. Ducasse E. Fully automatic volume segmentation of infrarenal abdominal aortic aneurysm computed tomography images with deep learning approaches versus physician controlled manual segmentation.J Vasc Surg. 2021; 74: 246-256 e6Abstract Full Text Full Text PDF Scopus (20) Google Scholar, 5Adam C. Fabre D. Mougin J. Zins M. Azarine A. Ardon R. et al.Pre-surgical and Post-surgical Aortic Aneurysm Maximum Diameter Measurement: Full Automation by Artificial Intelligence.Eur J Vasc Endovasc Surg. 2021; Abstract Full Text Full Text PDF Scopus (12) Google Scholar). These studies showed a good accuracy of the methods compared to human experts and demonstrated the feasibility to use AI for an automatic measurement of AAA maximal diameter (4Caradu C. Spampinato B. Vrancianu A.M. Berard X. Ducasse E. Fully automatic volume segmentation of infrarenal abdominal aortic aneurysm computed tomography images with deep learning approaches versus physician controlled manual segmentation.J Vasc Surg. 2021; 74: 246-256 e6Abstract Full Text Full Text PDF Scopus (20) Google Scholar, 5Adam C. Fabre D. Mougin J. Zins M. Azarine A. Ardon R. et al.Pre-surgical and Post-surgical Aortic Aneurysm Maximum Diameter Measurement: Full Automation by Artificial Intelligence.Eur J Vasc Endovasc Surg. 2021; Abstract Full Text Full Text PDF Scopus (12) Google Scholar, 6Lareyre F. Chaudhuri A. Flory V. Augene E. Adam C. Carrier M. et al.Automatic measurement of maximal diameter of abdominal aortic aneurysm on computed tomography angiography using artificial intelligence.Ann Vasc Surg. 2021; Google Scholar). Hence, CNN offers perspectives to develop applications oriented toward screening and identification of AAA as well as new tools to facilitate its anatomic characterization, which could improve pre-surgical planning and follow-up. As stated by the authors, in addition to develop advanced imaging analysis, ML has the potential to build predictive models of patients’ outcomes and several studies have underlined its interest to better assess AAA growth, risk of rupture, as well as risks of post-operative complications including mortality or re-interventions (7Raffort J. Adam C. Carrier M. Ballaith A. Coscas R. Jean-Baptiste E. et al.Artificial intelligence in abdominal aortic aneurysm.J Vasc Surg. 2020; Abstract Full Text Full Text PDF Scopus (53) Google Scholar, 8Raffort J. Adam C. Carrier M. Lareyre F. Fundamentals in Artificial Intelligence for Vascular Surgeons.Ann Vasc Surg. 2020; 65: 254-260Abstract Full Text Full Text PDF PubMed Scopus (25) Google Scholar, 9Lareyre F. Raffort J. Looking for the Optimal Evaluation of Abdominal Aortic Aneurysm Risk of Rupture.J Endovasc Ther. 2020; 27: 345-346Crossref Scopus (0) Google Scholar). AI has the potential to enhance precision medicine and although further studies are required to evaluate the accuracy and external validation, applications for clinical practice can hopefully be expected within the next few years. This work has been supported by the French government, through the 3IA Côte d’Azur Investments in the Future project managed by the National Research Agency (ANR) with the reference number ANR-19-P3IA-0002.