Perfusion Imaging Model Fitting for Artifact Removal
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current medical imaging techniques for blood perfusion analysis, particularly ultrasound imaging, face challenges such as poor image quality due to variations in pixel intensity over time, speckle grains, motion artifacts, and background echo signals, making it difficult to accurately assess perfusion processes.
Innovation Solution
A medical imaging system that processes a sequence of input images by associating each pixel or group of pixels with a model function over time, generating computed images with visualizing values determined by instantaneous function-values or integrals of the model functions, and displaying these as an animated sequence to enhance perfusion process visualization.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If a sequence of consecutive images is acquired to monitor contrast agent flow, then perfusion information is obtained, but image quality deteriorates due to large pixel intensity variations and speckle grains
Solution Approach 1:
The patent applies preliminary action by fitting a model function to the echo signal intensity over time before generating the final perfusion image. This model fitting process (Equation 1: I(t) = A·(1−e^−βt)) preprocesses the raw image data to extract meaningful perfusion parameters while filtering out noise and artifacts, thereby obtaining accurate perfusion information while maintaining high image quality
Solution Approach 2:
The patent transforms the raw image data by changing parameters from pixel intensity values to perfusion parameters (A and β). This parameter transformation converts the problematic raw echo signal variations into stable, meaningful perfusion characteristics that accurately represent blood flow while eliminating the visual noise of speckle grains and intensity fluctuations
2Measurement precision
If parametric analysis techniques are used to provide quantitative assessment, then perfusion parameters are calculated, but the perfusion process dynamics are lost in static representations
Solution Approach 1:
The patent adds the time dimension back to the parametric analysis by generating an animated sequence of perfusion images. Instead of displaying static parametric maps, the system creates a temporal sequence where each frame represents the perfusion state at different time points, allowing observers to visualize both the quantitative parameters and the dynamic evolution of the perfusion process simultaneously
3Object-affected harmful factors
If background echo signals are removed to enhance useful information, then signal-to-noise ratio improves, but quantitative accuracy may be affected
Solution Approach 1:
The patent applies local quality by selectively processing different components of the echo signal. The model fitting function is applied locally to each pixel or region of interest, extracting perfusion parameters from the contrast agent signal while inherently filtering out background echo signals. This localized processing preserves the quantitative accuracy of the perfusion measurement while removing harmful background interference
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
The system provides a clearer and more enhanced perception of perfusion dynamics by smoothing temporal and spatial intensities, removing motion artifacts, and suppressing background signals, facilitating better visualization and analysis of perfusion kinetics.
Implementation Method 1
The contrast agent acts as an efficient ultrasound reflector, so that it can be easily detected by applying ultrasound waves and recording a resulting echo signal
Data Source
AI summary
An embodiment of a medical imaging system is proposed. The system includes means for providing a sequence of recorded input images each one offering a digital representation at a corresponding instant of a body part being perfused with a contrast agent, each input image including a plurality of visualizing values each one representing a corresponding portion of the body part, and means for associating each sequence of corresponding sets in the input images of at least one visualizing value with a model function of time; the system further includes means for generating a sequence of computed images at further instants, each computed image including a plurality of further visualizing values each one being determined by an instantaneous function-value which is calculated from the associated model function at the corresponding further instant, and means for displaying the sequence of computed images.