Stenosis Localization via Device Feature Matching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Localizing stenoses in angiograms is complex due to the need for rigid protocols like cardiac roadmapping, which can be cumbersome and prone to inaccuracies, especially when vessel distortion occurs during treatment.
Innovation Solution
A device and method that utilize a data processor to identify and delineate stenoses in angiographic images based on device-related content from treatment X-ray images, employing a self-learning algorithm and convolutional network configuration to facilitate stenosis localization without registration, allowing for direct identification of structures similar to the interventional device.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If rigid cardiac roadmapping protocol is used for registration, then localization accuracy is improved, but device complexity and ease of operation deteriorate
Solution Approach 1:
The patent replaces the mechanical/manual registration process with an automated image processing system using machine learning algorithms. The system automatically detects the interventional device in treatment images, extracts its position and orientation, and localizes the stenosis in angiographic images without requiring manual protocol execution, thereby reducing complexity while maintaining accuracy
Solution Approach 2:
The system performs self-registration by automatically detecting device features and computing transformations without human intervention. The machine learning model autonomously identifies correspondences between treatment and angiographic images, eliminating the need for operators to follow complex roadmapping protocols
2Measurement precision
If rigid cardiac roadmapping protocol is used for registration, then localization accuracy is improved, but ease of operation deteriorates
Solution Approach 1:
The manual operation of following rigid protocols is replaced by an automated computational system. The machine learning model automatically processes images, detects devices, and performs localization, making the system as easy to operate as simply providing input images while achieving high accuracy
Solution Approach 2:
The system autonomously performs all registration and localization tasks without requiring operator expertise in executing complex protocols. The self-learning algorithm adapts to different cases automatically, making the system accessible and easy to operate for users with varying levels of expertise
3Adaptability or versatility
If registration is performed to handle vessel distortion, then localization accuracy deteriorates, but if no registration is used, then adaptability improves
Solution Approach 1:
The system uses a dynamic, flexible approach by employing machine learning models that can adapt to various vessel configurations and distortions. Rather than applying rigid registration transformations, the model learns to directly localize stenoses in angiographic images based on device features, accommodating vessel movement and distortion naturally
Solution Approach 2:
The system creates a direct mapping from device features in treatment images to stenosis locations in angiographic images without intermediate registration steps. This copying approach bypasses the registration problem entirely, maintaining accuracy while adapting to vessel distortion through the learning model's ability to generalize across different anatomical variations
Data Source
AI summary
The present invention relates to localizing stenoses. In order to provide improved and facilitated stenosis localization, a device (10) for localizing a stenosis in an angiogram is provided. The device comprises an image supply (12), a data processor (14) and an output (16). The image supply is configured to provide a first image (18) and a second image (20). The first image is an angiographic image that comprises image data representative of a region of interest of a vascular structure in a visible and distinct manner, wherein the vascular structure comprises at least one vessel with at least a part of a stenosis. The second image is a treatment X-ray image that comprises image data representative of at least a part of an interventional device arranged within the vascular structure in a state when the stenosis of the vascular structure is treated. The data processor is configured to identify and delineate the stenosis in the first image based on the first image and at least based on device-related content present in the second image. The data processor is also configured to detect the interventional device in the second image, and to provide a direct identification of structures in the first image that are most similar to the device as detected in the second image. The output is configured to provide an indication of the stenosis.