Volumetric Angiographic Image Correction for Clearer Vessel Trees
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Volumetric angiographic images suffer from limitations in image quality, such as inaccurate vessel diameters, poor visualization of thin fibrous caps and thrombi, and image artifacts like blooming, which hinder reliable vascular diagnoses, while intravascular images are invasive and time-consuming.
Innovation Solution
A system using a machine learning algorithm trained on vessel image pairs to generate corrected volumetric angiographic images with improved image quality, eliminating the need for invasive intravascular imaging.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If volumetric angiographic imaging is used to provide comprehensive vascular information, then diagnostic coverage is improved, but image quality and measurement precision deteriorate
Solution Approach 1:
The patent combines volumetric angiographic imaging (providing comprehensive vascular coverage) with intravascular imaging (providing high image quality) by integrating multiple imaging modalities into a single system. The imaging catheter incorporates both angiographic imaging capabilities and intravascular imaging capabilities, allowing simultaneous or sequential acquisition of both types of images during a single procedure, thereby resolving the contradiction between diagnostic coverage and image quality
Solution Approach 2:
The patent introduces an image fusion algorithm as an intermediary that processes and integrates data from both volumetric angiographic images and intravascular images. This algorithm fuses the complementary information from both imaging modalities to generate enhanced diagnostic images that maintain the comprehensive vascular coverage of angiography while incorporating the high image quality and detailed vessel wall information from intravascular imaging
2Measurement precision
If intravascular imaging is used to improve image quality, then measurement precision is improved, but procedural time and complexity increase
Solution Approach 1:
The patent merges angiographic imaging and intravascular imaging into a single integrated procedure using a combined imaging catheter. Both imaging modalities are performed during one continuous procedure rather than requiring separate procedures, thereby reducing total procedural time while maintaining the high image quality benefits of intravascular imaging
Solution Approach 2:
The patent performs volumetric angiographic imaging before intravascular imaging to identify vessels requiring detailed assessment. This preliminary screening allows the operator to target only specific vessels for intravascular imaging rather than imaging all vessels, thereby reducing the time required for intravascular imaging while still achieving high image quality where needed
3Adaptability or versatility
If intravascular imaging is performed in multiple vessel branches, then diagnostic coverage is improved, but procedural complexity and time increase
Solution Approach 1:
The patent employs a universal imaging catheter that can perform both volumetric angiographic imaging and intravascular imaging functions. This multi-functional device can assess multiple vessel branches using the same catheter and procedure, eliminating the need for different imaging approaches for different vessel types and thereby reducing procedural complexity while maintaining comprehensive diagnostic coverage
Data Source
Figure 1~2
Figure 3
AI summary
A system for improving image quality in volumetric angiographic images, is provided. The system includes one or more processors configured to input a volumetric angiographic image (120) representing a vessel tree into a machine learning algorithm (130). The machine learning algorithm is trained to generate, based on the inputted volumetric angiographic image, a corrected volumetric angiographic image (140) wherein an image quality of the vessel tree is improved as compared to in the inputted volumetric angiographic image. The one or more processors are also configured to output the corrected volumetric angiographic image (140). The machine learning algorithm (130) is trained to generate the corrected volumetric angiographic image (140) using training data (150) comprising a plurality of vessel image pairs, each vessel image pair comprising a volumetric angiographic image (150a) representing a vessel and a corresponding volumetric intravascular image (150b) representing the vessel.