Vascular Image Co-Registration for Hemodynamic Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing medical imaging systems and methods for vascular imaging lack effective co-registration of intravascular and extravascular data, limiting the integration and display of complementary information for accurate hemodynamic assessment.
Innovation Solution
A neural network is trained using co-registered intravascular and extravascular imaging data sets, including intravascular ultrasound, optical coherence tomography, and angiographic data, to predict hemodynamic values from intravascular images, enabling simultaneous display of both types of data for enhanced vascular assessment.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If intravascular and extravascular imaging data are co-registered using traditional methods, then integration of imaging data is achieved, but additional pullback runs are required which reduce productivity
Solution Approach 1:
The system performs preliminary co-registration by capturing extravascular imaging data (angiography, CTA, or MRA) before the intravascular pullback run. This pre-acquired extravascular data is then co-registered with the intravascular images during the single pullback run, eliminating the need for additional pullback runs and improving productivity while maintaining complete data integration
Solution Approach 2:
The system creates a virtual copy of the extravascular imaging data and co-registers it with the intravascular images. This virtual co-registration allows the display system to present integrated intravascular and extravascular information without physically performing additional imaging runs, thereby maintaining information completeness while improving diagnostic efficiency
2Measurement precision
If multiple imaging modalities are integrated for comprehensive vascular assessment, then diagnostic accuracy is improved, but system complexity increases
Solution Approach 1:
The display system is designed with multi-functionality to handle multiple imaging modalities (intravascular ultrasound, optical coherence tomography, angiography, CTA, and MRA). The system can selectively display different combinations of imaging data based on clinical needs, providing comprehensive vascular assessment through a single unified interface without requiring separate systems for each modality
Solution Approach 2:
The system uses co-registration technology as an intermediary process that bridges intravascular and extravascular imaging data. This mediator aligns the different imaging modalities in a common coordinate system, allowing accurate spatial correlation and integrated display without requiring direct complex integration between all imaging systems
3Measurement precision
If hemodynamic values are obtained through additional pullback runs, then measurement accuracy is improved, but procedure time increases
Solution Approach 1:
The system performs preliminary acquisition of extravascular imaging data and co-registers it with intravascular images during the single pullback run. This preliminary preparation enables the system to derive hemodynamic information from the co-registered images without requiring additional pullback runs, thereby maintaining measurement accuracy while reducing procedure time
Solution Approach 2:
The system replaces the mechanical approach of performing additional physical pullback runs with a computational approach. By using image processing and co-registration algorithms on already-acquired intravascular and extravascular images, the system derives hemodynamic values without the need for repeated mechanical catheter movements, thus saving time while maintaining accuracy
Data Source
AI summary
A neural network is trained for estimating patient hemodynamic data using a plurality of extravascular imaging data sets and a plurality of intravascular imaging data sets that are each co-registered to a corresponding extravascular imaging data set. A plurality of hemodynamic data sets are provided, each hemodynamic data set co-registered with the corresponding extravascular imaging data set. The neural network learns what hemodynamic data to expect for a given intravascular imaging data set. An intravascular imaging event is subsequently performed in which an intravascular imaging element is translated within a blood vessel of the patient to produce one or more intravascular images. The neural network uses its training to predict hemodynamic values corresponding to the one or more intravascular images from the intravascular imaging event, and the one or more intravascular images are outputted in combination with the predicted hemodynamic values.