Arterial Input Function Determination via Recirculation Kernel Modeling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current indicator-dilution techniques in medical imaging, such as PET, CT, and MRI, fail to accurately account for tracer recirculation, leading to biases in estimating local blood volume and arterial input function, particularly due to the assumption of independent input and output in tracer transport models.
Innovation Solution
A system and method that models blood flow in the global circulatory system using multiple delayed or dispersed input response functions to determine the arterial input function and absolute blood volume, incorporating recirculation effects by integrating tracer concentrations over time, allowing for the estimation of blood volume without isolating the first bolus passage.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the first passage of tracer through the organ is extracted to conform to open pipeline models, then the measured tracer amount can be fitted to tracer kinetic models, but information is lost and biases are introduced in estimating local blood volume
Solution Approach 1:
The patent extracts the recirculation component from the total tracer concentration curve by modeling it as a separate convolved function of the arterial input function and a recirculation kernel. This allows the first passage (extraction) to be performed without discarding recirculation information, as the recirculation is explicitly modeled and removed mathematically rather than by ad hoc cutoff, thereby reducing information loss and bias in blood volume estimation.
Solution Approach 2:
The patent changes the parameter representation by introducing a recirculation kernel function and using convolution operations to model recirculation dynamics. This transforms the problem from simply cutting off recirculation data to mathematically characterizing and separating recirculation based on its temporal and concentration parameters, improving the accuracy of extracted first passage data.
2Ease of manufacture
If an ad hoc cut-off is used to find the area under the single bolus, then the calculation can be performed, but biases are introduced in the estimation of local blood volume
Solution Approach 1:
The patent introduces a recirculation kernel function as an intermediary mathematical construct that mediates between the total tracer curve and the first passage extraction. Instead of using ad hoc cut-off methods, the recirculation kernel serves as a mathematical mediator that explicitly models recirculation dynamics, allowing for more accurate separation of first passage and recirculation components while maintaining computational feasibility through convolution operations.
3Device complexity
If tracer recirculation is not accounted for in tracer transport models, then the models can assume independent input and output, but the models fail to accurately represent blood circulation
Solution Approach 1:
The patent segments the tracer transport process into distinct components: the arterial input function (first passage), the recirculation component, and their convolution. By dividing the total tracer concentration curve into these separable components with different temporal characteristics, the model can account for recirculation effects without requiring complete redesign of the tracer transport framework, thus improving reliability while controlling complexity.
Solution Approach 2:
The patent maintains continuous modeling of tracer dynamics by using convolution operations that inherently account for the continuous nature of recirculation. Rather than interrupting the tracer curve with ad hoc cut-offs, the convolution-based approach continuously models the叠加 of first passage and recirculation boluses over time, providing a more reliable and continuous representation of blood circulation dynamics.
Data Source
AI summary
An exemplary system, method and computer-accessible medium for determining an arterial input function (AIF) of a mammal(s) can be provided, which can include, for example, receiving information related to a global circulatory system of the mammal(s), and determining the AIF based on the information by modeling a blood flow in the global circulatory system of the mammal(s) in terms of an input response function(s). The input response function(s) can include a delayed input response function(s). In certain exemplary embodiments of the present disclosure, the input response function(s) can include at least three input response functions, and each of the input response functions can be from a different part of a body of the mammal(s). The AIF can be determined by coupling the input response functions. The AIF can be further determined based on a total tracer amount in an organ(s) of the mammal(s).