Vascular Function Extraction from Brain CT Using Deep Learning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for analyzing blood flow, such as phase-contrast magnetic resonance imaging, are insufficient in providing high temporal resolution and accurately deriving artery and vein functions, leading to errors in blood flow analysis due to inaccurate artery function calculations.
Innovation Solution
An apparatus and method that extract vascular functions from brain-related information using a CT image, applying time interpolation, deep learning-based vessel segmentation, and motion correction to generate accurate artery and vein functions, enabling precise calculation of blood flow parameters like CBF, CBV, and MTT.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If phase-contrast magnetic resonance imaging is used to measure blood flow, then non-invasive measurement is achieved, but temporal resolution and measurement precision are insufficient
Solution Approach 1:
The patent changes the imaging modality from MRI to CT, and transforms the static anatomical images into dynamic functional images by introducing time as a parameter. The CT perfusion imaging captures multiple time points to create temporal resolution, while the deconvolution algorithm transforms the concentration-time curves into physiological parameters (CBF, CBV, MTT), thereby achieving both temporal resolution and measurement precision simultaneously.
2Reliability
If generalization and Gaussian transform methods are used to derive artery and vein functions, then some vascular functions can be obtained, but accuracy deteriorates due to errors in artery function derivation
Solution Approach 1:
The patent extracts the artery function derivation from the conventional generalization and Gaussian transform methods, and replaces it with a dedicated deconvolution algorithm specifically designed for this purpose. By separating this critical function and applying a more appropriate mathematical method, the patent achieves higher precision in artery function derivation, which then improves the overall reliability of vascular function assessment.
Solution Approach 2:
The patent substitutes the mechanical/mathematical approach of generalization and Gaussian transform with a deconvolution algorithm that is better suited for extracting arterial input functions from CT perfusion data. This mathematical substitution provides more accurate results by directly solving the convolution equation that describes contrast agent dynamics in the vasculature.
3Measurement precision
If conventional methods are used to analyze blood flow, then basic velocity information can be obtained, but accuracy of blood flow analysis deteriorates
Solution Approach 1:
The patent transitions from one-dimensional velocity measurement to multi-dimensional physiological parameter analysis. By introducing time as an additional dimension and applying deconvolution to the concentration-time curves, the patent extracts multiple physiological parameters (CBF, CBV, MTT) that provide a comprehensive view of blood flow dynamics, thereby achieving both higher accuracy and more complete information.
Data Source
AI summary
An aspect of the present disclosure provides an apparatus for extracting a vascular function including an information reception unit configured to extract an original CT image from brain-related information received from the outside; an NIFTI image transformation unit configured to transform the original CT image into an NIFTI file format image to acquire time sequence data; a time interpolation unit configured to apply time interpolation to the original CT image through the time sequence data to transform the original CT image into each time-specific 3D CT image; a vessel segmentation unit configured to predict a vessel segmentation mask by passing the each time-specific 3D CT image through a deep learning-based vessel segmentation deep-learning model 141 and generate a 4D vessel mask image by stacking the 3D CT images based on a time axis; and a vascular function extraction unit configured to extract a vascular function from a vessel region of the 4D vessel mask image and calculate a blood flow parameter using an artery function which is one of the vascular functions.