Deep learning method for aortic root detection

• Published in: Computers in biology and medicine Vol. 135; p. 104533
• Main Authors: Tahoces, Pablo G., Varela, Rafael, Carreira, Jose M.
• Format: Journal Article
• Language: English
• Published: Oxford Elsevier Ltd 01-08-2021; Elsevier Limited
• Subjects: Abdomen; Algorithms; Aneurysms; Angiography; Aorta; Aortic root; Automation; Computed tomography; Computed tomography angiography (CTA); Coronary vessels; Datasets; Deep learning; Detection; Disease; Feasibility studies; Heart; Image processing; Image segmentation; Imaging techniques; Landmarks; Morphology; Test sets; Vascular diseases; Vascular imaging; Aortic root; Landmarks; Computed tomography angiography (CTA); Detection; Vascular imaging
• ISSN: 0010-4825; 1879-0534
• DOI: 10.1016/j.compbiomed.2021.104533

Computed tomography angiography (CTA) is a preferred imaging technique for a wide range of vascular diseases. However, extensive manual analysis is required to detect and identify several anatomical landmarks for clinical application. This study demonstrates the feasibility of a fully automatic method for detecting the aortic root, which is a key anatomical landmark in this type of procedure. The approach is based on the use of deep learning techniques that attempt to mimic expert behavior.
A total of 69 CTA scans (39 for training and 30 for validation) with different pathology types were selected to train the network. Furthermore, a total of 71 CTA scans were selected independently and applied as the test set to assess their performance.
The accuracy was evaluated by comparing the locations marked by the method with benchmark locations (which were manually marked by two experts). The interobserver error was 4.6 ± 2.3 mm. On an average, the differences between the locations marked by the two experts and those detected by the computer were 6.6 ± 3.0 mm and 6.8 ± 3.3 mm, respectively, when calculated using the test set.
From an analysis of these results, we can conclude that the proposed method based on pre-trained CNN models can accurately detect the aortic root in CTA images without prior segmentation.
[Display omitted]
•The development of new algorithms that provide quantitative characteristics from CT images has become an important issue.•The aortic root is an essential reference point to extract quantitative information on the shape and size of the aorta.•The approach allows detecting the aortic root for non-ECG gated CTA cases and no previous segmentation of the aorta.