X-Ray Vessel Flow Quantification Using Spatio-Temporal Frequency Maps
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current medical imaging techniques, such as 2D and 4D DSA, lack the ability to accurately quantify blood flow and velocity in complex 3D vascular structures, leading to subjective and unreliable clinical assessments during procedures like liver embolization, with existing methods being sensitive to noise, bolus dispersion, and requiring iterative processing.
Innovation Solution
A system and method for generating quantitative flow or velocity information using a single-step algorithm in the spatiotemporal frequency domain, utilizing x-ray data to create a time-attenuation map and perform a frequency domain transform to identify peak frequencies, allowing for non-iterative velocity quantification in vessels.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional 2D DSA or iterative qDSA methods are used to extract velocity information, then velocity can be estimated, but the measurements are highly sensitive to image noise and bolus dispersion, reducing accuracy
Solution Approach 1:
The patent transforms the problem from temporal analysis of TACs to spatio-temporal frequency domain analysis. By applying 2D Fourier transform to the time-attenuation map (adding spatial frequency dimension to temporal frequency), the method captures velocity information in the frequency domain, making the measurement robust against noise and bolus dispersion effects that plague conventional temporal analysis methods
Solution Approach 2:
The patent replaces iterative mechanical optimization processes with a direct frequency domain transformation approach. Instead of iteratively adjusting parameters to minimize error between TACs, the method directly computes velocity from the spatio-temporal frequency spectrum, eliminating sensitivity to initialization and convergence issues while reducing computational complexity
2Measurement precision
If iterative least-squares or cross-correlation methods are used to align TACs, then velocity can be calculated, but the iterative processing increases computational complexity and processing time
Solution Approach 1:
The patent substitutes iterative optimization algorithms with a direct Fourier transform-based frequency domain method. The velocity is obtained directly from the spatio-temporal frequency spectrum without requiring iterative alignment or parameter optimization, dramatically simplifying the computational process while maintaining accuracy
Solution Approach 2:
The patent changes the domain of analysis from time-domain TAC comparison to frequency-domain spatio-temporal spectral analysis. This parameter transformation allows velocity to be extracted directly from frequency relationships rather than through iterative temporal alignment, reducing computational complexity
3Loss of information
If 4D DSA rotational acquisition is performed to obtain 3D vessel anatomy, then complete vascular structure information is available, but the acquisition time and radiation exposure double due to requiring two separate rotational sweeps
Solution Approach 1:
The patent extracts only the essential information needed for velocity quantification from the x-ray data - the time-attenuation characteristics along vessel centerlines. By focusing on extracting velocity information directly from contrast dynamics rather than acquiring complete 3D anatomical datasets, the method achieves velocity measurement with a single rotational sweep, halving acquisition time and radiation exposure
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Provides accurate and consistent velocity quantification in the presence of pulsatile flow, reducing sensitivity to noise and motion artifacts, and enabling reliable clinical decision-making without iterative processing.
Implementation Method 1
an x-ray imaging system configured to acquire x-ray data from the subject while a contrast agent is delivered to the vessel of the subject
Data Source
AI summary
A system and method are provided that includes acquiring or accessing x-ray data of a subject experiencing a delivery of a contrast agent to the vessel of the subject and generating time-attenuation curves for the vessel using the x-ray data. The method also includes identifying a plurality of points within the vessel and extending along a lumen of the vessel and sampling the time-attenuation curves at the plurality of points to generate a time-attenuation map. The method further includes performing a Fourier transform on the time-attenuation map to generate a spatio-temporal map of spatial frequency versus temporal frequency and identifying a peak frequency in the spatio-temporal map corresponding to flow or velocity in the vessel. The method then includes quantifying the flow or velocity within the vessel of the subject using the peak frequency and generating a report including quantified flow or velocity through the vessel.