3D Blood Vessel Segmentation for Aneurysm Diameter Measurement
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for diagnosing and monitoring cardiovascular diseases, particularly abdominal aortic aneurysms, are manual, time-consuming, and practitioner-dependent, leading to inconsistent and unreliable measurements of geometric indicators like diameter, which are crucial for decision-making on treatment and post-treatment monitoring.
Innovation Solution
A method using a three-dimensional representation of a blood vessel segmented by a classifier to automatically identify voxels belonging to the vessel, allowing for reliable determination of geometric indicators like diameter, involving machine learning algorithms and convolutional neural networks to distinguish between vessel lumen, vessel walls, and potential obstructions like stents or calcifications.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual examination of angiograms is used to detect aneurysms, then the practitioner can identify significant local variations in aortic diameter, but the process becomes time-consuming and practitioner-dependent with inconsistent measurements
Solution Approach 1:
The patent replaces the manual mechanical examination process with an automated computer-based system that processes angiogram images. The system automatically detects aneurysms by analyzing local variations in aortic diameter through computational algorithms, eliminating the need for manual practitioner examination while improving measurement consistency and reducing time requirements.
Solution Approach 2:
The system enables self-service by allowing the computer to autonomously perform the entire diagnostic process without practitioner intervention. The automated system independently examines angiograms, detects aneurysms, measures diameters, and generates diagnostic information, making the process independent of practitioner availability and expertise levels.
2Reliability
If manual diameter calculation is performed by selecting specific images, then the practitioner can determine lumen diameter, but the results become dependent on practitioner expertise and not reproducible
Solution Approach 1:
The patent replaces manual image selection and diameter measurement with an automated classification system that processes all angiogram images systematically. The computer-based system applies consistent algorithms to identify and measure aneurysms across multiple images, ensuring reproducible results independent of practitioner expertise while managing complexity through standardized computational procedures.
Solution Approach 2:
The automated classification system serves multiple functions: it selects relevant images, detects aneurysms, measures diameters, and generates diagnostic reports. This multi-functional system replaces multiple manual steps with a single integrated computational process, improving reliability while the standardized approach manages system complexity.
3Measurement precision
If stents or calcifications are present in the aorta, then blood flow is affected, but manual methods cannot reliably distinguish these elements from the vessel lumen
Solution Approach 1:
The patent applies segmentation by dividing the aortic structure into distinct components: vessel lumen, stents, and calcifications. The classification system processes angiogram images to identify and separate these different elements based on their visual characteristics, allowing precise measurement of the true lumen diameter while accounting for the presence of stents or calcific deposits that would otherwise be misidentified.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The invention relates to a method for aiding in the diagnosis of a cardiovascular disease (A), comprising the following steps: (E0) providing a three-dimensional representation of (1) a blood vessel (AA) of a patient; (E1) segmenting, by means of a classifier (CNN), the three-dimensional representation to obtain a segmented three-dimensional map (2); (E3) comparing the value of a plurality of voxels of the three-dimensional representation, the allocated labels on the three-dimensional map of the voxels being those of the blood vessel, with a predetermined threshold value, a label different from those of the blood vessel being allocated to each voxel with a value that exceeds the predetermined threshold value; (E4) determining the change in a geometric indicator (Di, Dj) of the blood vessel by means of the voxels of the three-dimensional representation, the allocated labels on the three-dimensional map of the aforementioned voxels being those of the blood vessel.