IVUS Vessel Segment Identification Without X-Ray Annotation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for identifying blood vessel segments in intraluminal imaging, such as X-ray and IVUS, are tedious and error-prone, requiring manual annotation and exposing clinicians to radiation, which is harmful.
Innovation Solution
An intraluminal imaging system using computer vision algorithms and probabilistic models, like Bayesian graph networks, automatically identifies blood vessel segments from IVUS images, reducing the need for X-ray images and manual identification.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If manual annotation and X-ray imaging are used to identify blood vessel segments, then identification accuracy can be achieved, but the process becomes tedious and exposes clinicians to radiation
Solution Approach 1:
The patent replaces manual annotation and X-ray imaging with an automated computer vision system that uses intraluminal ultrasound images to identify blood vessel segments. The system processes ultrasound images through algorithms to automatically determine vessel segment locations, eliminating the need for radiation-based X-ray imaging and manual clinician annotation while maintaining identification accuracy.
2Reliability
If manual annotation is used to label vessel segments, then identification can be performed, but it increases the time burden on clinicians
Solution Approach 1:
The system enables self-service automation where the computer vision algorithm independently processes intraluminal ultrasound images to identify and label vessel segments without requiring clinician intervention. The automated system performs the entire annotation process that would otherwise require manual clinician time, significantly reducing the time burden while maintaining accurate identification.
3Loss of information
If X-ray imaging is used to locate and assess lesion sites, then diagnostic information can be obtained, but radiation exposure increases
Solution Approach 1:
The patent substitutes radiation-based X-ray imaging with non-radiation intraluminal ultrasound imaging to obtain diagnostic information about lesion sites and blood vessel segments. The system extracts diagnostic information directly from ultrasound images through computer vision algorithms, eliminating the need for additional radiation exposure while maintaining comprehensive diagnostic capability.
Data Source
AI summary
An intraluminal ultrasound imaging system is provided, which includes a processor in communication with an intraluminal ultrasound imaging catheter. The processor is configured to receive an intraluminal ultrasound image associated with a first segment of the body lumen from which the image was captured, which includes depictions of the body lumen and a second body lumen. The processor receives a second image obtained while the intraluminal ultrasound imaging catheter is moving at a pullback speed within the body lumen. The second image also includes depictions of the body lumen and the second body lumen. The processor computes body lumen properties from each image, and determines a second segment of the body lumen for the second image, based on a model associated with the body lumen properties, and/or the pullback speed, and outputs an indication of the second segment to a display.