Abstract:Introduction: Contrast medium (CM) use in computed tomography (CT) is limited by nephrotoxicity and possible allergic reactions. The purpose of this study is to introduce a tool for the diagnosis of abdominal aortic aneurysms (AAAs) by avoiding the use of CM. Methods: With and without CM CTs of patients with AAA were evaluated. A mathematical algorithm was implemented to allow visualization of the inner aortic lumen in the series without CM. The first step of the algorithm consisted in manually highlighting a squared region of interest (ROI) close to the target aortic area. The rest of the algorithm is completely automated and performs the following flow of operations: The "Kantorovich" algorithm is applied to the ROI for image enhancement. Then, a wavelet decomposition method is applied to identify the different frequency components of the image. Exploiting the wavelet decomposition, the system selects the low-frequency components of the image, corresponding to the major structures. Thresholding method, based on the analysis of the gray-level histogram, is then performed to extract the contours of the vessel. At this point, the extraction of the pervious area is completed. Final images were compared with the contrast enhanced scans, valued as gold standard. To validate the algorithm, an analysis of the results has been performed considering the following types of error: En = #m/#ROI ΔA = |1-(#CM/#CEX)| (#m = number of misclassified pixels; #ROI = number of pixels in the ROI; #CM = number of pixels belonging to the CM; #CEX = extracted areas). En provided a measure on the pixels wrongly classified, and ΔA the difference of extracted areas. Results: The algorithm was applied to 233 CT images. Extracted images were compared with the ones with CM. Mean values of the numerical errors ranged from 0.12 to 0.17 for En and 0.11 to 0.35 for ΔA. For all the 233 CT images in the analyzed sequences, the mean error was 0.14 (En) and 0.28 (ΔA). Conclusions: The developed mathematical algorithm allows to detect the position of the vessel lumen and to extract its contours with a good accuracy. Our experience shows encouraging results and suggests a possible future clinical application to reduce CT CM use.

CACTUSS: Common Anatomical CT-US Space for US examinations

Segmentation of Cardiac Tissues and Organs for CCTA Images Based on a Deep Learning Model

Deep learning-based segmentation of abdominal aortic aneurysms and intraluminal thrombus in 3D ultrasound images

Computed Tomography Postprocessing for Abdominal Aortic Aneurysm Lumen Recognition in Unenhanced Examinations

Multi-Task Deep Convolutional Neural Network for the Segmentation of Type B Aortic Dissection

Automatic Segmentation of Thoracic Aorta Segments in Low-Dose Chest CT

Automatic Segmentation of Coronary Lumen and External Elastic Membrane in Intravascular Ultrasound Images Using 8-Layer U-Net

A Deep Learning Approach to Automate High-Resolution Blood Vessel Reconstruction on Computerized Tomography Images With or Without the Use of Contrast Agent

Fully Automatic Segmentation of Abdominal Aortic Thrombus in Pre-Operative CTA Images Using Deep Convolutional Neural Networks.

Automated Screening for Abdominal Aortic Aneurysm in CT Scans under Clinical Conditions Using Deep Learning

Computed tomography-based automated measurement of abdominal aortic aneurysm using semantic segmentation with active learning

Detection and Visualization of Endoleaks in CT Data for Monitoring of Thoracic and Abdominal Aortic Aneurysm Stents

DeepVox and SAVE-CT: a contrast- and dose-independent 3D deep learning approach for thoracic aorta segmentation and aneurysm prediction using computed tomography scans

Segmentation of Aorta 3D CT Images Based on 2D Convolutional Neural Networks

Automatic Segmentation of Type A Aortic Dissection on Computed Tomography Images Using Deep Learning Approach

A retrospective study of 3D deep learning approach incorporating coordinate information to improve the segmentation of pre- and post-operative abdominal aortic aneurysm

DeepAAA: clinically applicable and generalizable detection of abdominal aortic aneurysm using deep learning

Segment aorta and localize landmarks simultaneously on noncontrast CT using a multitask learning framework for patients without severe vascular disease

Deep-learning-based Image Segmentation for Image-based Computational Hemodynamic Analysis of Abdominal Aortic Aneurysms: A Comparison Study

CathFlow: Self-Supervised Segmentation of Catheters in Interventional Ultrasound Using Optical Flow and Transformers

Abdominal Aortic Aneurysm Segmentation with a Small Number of Training Subjects