Photon computed tomography pure calcification sequence analysis system for coronary artery plaque and stenosis

Abstract

The application discloses a coronary artery plaque and stenosis analysis system based on a photon CT pure calcification sequence and belongs to the technical field of coronary arteries. The system comprises a data acquisition sequence generation module, a calcification artifact precision correction module and a plaque and stenosis analysis and diagnosis module. The data acquisition sequence generation module is used for coronary artery scanning based on photon counting CT to obtain photon CT pure calcification sequence image data. The calcification artifact precision correction module is used for the precise correction of coronary artery calcification artifacts of CCTA images to determine a real lumen view without artifact interference. The plaque and stenosis analysis and diagnosis module is used for the analysis and diagnosis of coronary artery plaques and stenosis. The application solves the problem that the existing coronary artery plaques and stenosis cannot be accurately analyzed based on a photon CT pure calcification sequence, thereby leading to low analysis and diagnosis accuracy of coronary artery plaques and stenosis. The application can accurately analyze coronary artery plaques and stenosis based on a photon CT pure calcification sequence and can improve the analysis and diagnosis accuracy of coronary artery plaques and stenosis.

Description

Technical Field

[0001] This invention relates to the field of coronary artery technology, specifically to a system for analyzing coronary artery plaques and stenosis using pure calcification sequences obtained from photon CT. Background Technology

[0002] The coronary arteries are the arteries that supply blood to the heart. They are named for their crown-like shape around the heart. They originate from the root of the aorta and divide into left and right branches that run parallel to each other on the surface of the heart. Coronary angiography is the gold standard for diagnosing vascular stenosis and can clearly identify the location and extent of the lesion.

[0003] Coronary artery plaque with stenosis is a pathological condition in which plaques formed by atherosclerosis in the coronary arteries narrow the lumen of the blood vessels. It may cause myocardial ischemia, angina pectoris, or even myocardial infarction. Therefore, accurate analysis of coronary artery plaques and stenosis is particularly important.

[0004] Current technology cannot accurately analyze coronary artery plaques and stenosis based on pure calcification sequences from photon CT, resulting in low accuracy in the analysis and diagnosis of coronary artery plaques and stenosis. Summary of the Invention

[0005] The purpose of this invention is to provide an analysis system for coronary artery plaques and stenosis using pure calcification sequences from photonic CT. This system can accurately analyze coronary artery plaques and stenosis based on pure calcification sequences from photonic CT, thereby improving the accuracy of analysis and diagnosis of coronary artery plaques and stenosis and solving the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides the following technical solution:

[0007] A system for analyzing coronary artery plaques and stenosis using pure calcification sequences from photon CT, including:

[0008] The data acquisition sequence generation module is used to perform coronary artery scanning based on photon counting CT and acquire pure calcification sequence image data from photon CT.

[0009] The calcification artifact precision correction module is used to process pure calcification sequence image data from photon CT and construct a coronary artery calcification artifact correction model. It performs precise correction of coronary artery calcification artifacts on CCTA images to determine the true lumen view without artifact interference.

[0010] The plaque and stenosis analysis and diagnosis module is used to analyze and diagnose coronary artery plaques and stenosis based on real lumen views, accurately assessing the status of coronary artery plaques and stenosis.

[0011] Preferably, coronary artery scanning is performed based on photon counting CT to acquire pure calcification sequence image data from photon CT, and the following operations are performed:

[0012] Based on semiconductor detectors, incident X-ray photons are converted into electrical signals. By utilizing the differences in X-ray absorption characteristics of calcification, blood, and soft tissue at different energies, and leveraging the energy resolution capability of photon counting CT, calcified tissue is directly separated and quantified from the raw data level. Signals that conform to the energy characteristics of calcification are found and retained, while signals that do not conform to the energy characteristics of calcification are suppressed, generating a pure calcification sequence. This allows the acquisition of pure calcification sequence image data from photon CT, displaying only calcification information without interference from lumens and soft tissues.

[0013] Preferably, the pure calcification sequence image data from photon CT is processed by performing the following operations:

[0014] The pure calcification sequence image data of photon CT is cleaned to remove noise, and missing and outlier values ​​in the pure calcification sequence image data of photon CT are identified and evaluated.

[0015] Among them, it is necessary to determine whether missing and outlier values ​​in pure calcification sequence image data of photon CT are valuable for the analysis of coronary artery plaques and stenosis, and to process the missing and outlier values ​​in pure calcification sequence image data of photon CT according to the determination.

[0016] When missing and outlier values ​​in pure calcification sequence image data of photon CT are valuable for the analysis of coronary artery plaques and stenosis, the median is used to fill the missing values ​​in the pure calcification sequence image data of photon CT, and the mean is used to replace the outlier values ​​in the pure calcification sequence image data of photon CT.

[0017] When missing and outlier values ​​in the pure calcification sequence image data of photon CT are of no value for the analysis of coronary artery plaques and stenosis, then the missing and outlier values ​​in the pure calcification sequence image data of photon CT are removed.

[0018] Preferably, the process of processing the pure calcification sequence image data from photon CT also includes the following operations:

[0019] Enhancement processing is performed on pure calcification sequence image data from photon CT to emphasize the overall or local characteristics of the image, make the originally unclear image clearer, amplify the differences between object features in the image, and preserve the sharp edges of coronary artery calcification plaques;

[0020] Feature extraction was performed on pure calcification sequence image data from photon CT, and features related to coronary artery plaque and stenosis analysis were extracted from the pure calcification sequence image data from photon CT to determine the feature data of pure calcification sequence images from photon CT.

[0021] Preferably, to accurately correct coronary artery calcification artifacts in CCTA images, the following operations are performed:

[0022] A coronary artery calcification artifact correction model was constructed based on the feature data of pure calcification sequence images from photon CT. CCTA images were then input into the coronary artery calcification artifact correction model. The CCTA images were analyzed and identified according to the coronary artery calcification artifact correction model, and the coronary artery calcification artifacts were accurately corrected to determine the true lumen view without artifact interference, clearly displaying the real vascular lumen.

[0023] Preferably, a coronary artery calcification artifact correction model is constructed, and the following operations are performed:

[0024] The feature data of pure calcification sequence images in photon CT are divided into training set and test set.

[0025] The machine learning model is trained using a training set, enabling it to autonomously learn the coronary artery calcification artifact correction behavior from the training set and accurately correct coronary artery calcification artifacts, thus determining the coronary artery calcification artifact correction model.

[0026] The coronary artery calcification artifact correction model was tested using a test set, and its generalization performance was evaluated to determine the model test evaluation results.

[0027] The coronary artery calcification artifact correction model was optimized based on the model test evaluation results to determine the optimal coronary artery calcification artifact correction model.

[0028] Preferably, to evaluate the generalization performance of the coronary artery calcification artifact correction model, the following operations are performed:

[0029] The test set was input into the coronary artery calcification artifact correction model. The generalization performance of the coronary artery calcification artifact correction model was evaluated based on accuracy and F1 score, and it was determined whether the coronary artery calcification artifact correction model could achieve the expected effect of accurately correcting coronary artery calcification artifacts.

[0030] When the coronary artery calcification artifact correction model fails to achieve the expected effect of accurately correcting coronary artery calcification artifacts, the parameters of the coronary artery calcification artifact correction model are adjusted and iteratively optimized until the coronary artery calcification artifact correction model can achieve the expected effect of accurately correcting coronary artery calcification artifacts, and the optimal coronary artery calcification artifact correction model is determined.

[0031] Preferably, the analysis of coronary artery plaques and stenosis is performed based on a true lumen view free from artifact interference, to assess the degree of stenosis, and to quantitatively analyze non-calcified plaques and low-density plaques, providing a more comprehensive plaque burden assessment and determining the analysis and assessment report of coronary artery plaques and stenosis.

[0032] Preferably, the system for analyzing coronary artery plaques and stenosis using pure calcification sequences from photon CT is characterized by further including a plaque burden quantitative assessment module for performing the following steps:

[0033] Geometric and physical characteristics of coronary artery plaques were extracted from artifact-free, realistic lumen views, including plaque volume, average plaque density, luminal stenosis, plaque surface roughness, and blood flow velocity. Plaque volume was calculated using 3D reconstruction of the plaque region in the realistic lumen view, expressed in millimeters per cubic millimeter (mm³). Average plaque density was determined by calculating the average grayscale value of the plaque region in clean calcified CT images, expressed in Hounsfield units (HU). Lumen stenosis was calculated as a percentage (%) of the normal lumen diameter compared to the minimum diameter at the stenosis. Plaque surface roughness was determined by analyzing the pixel intensity variation rate at the plaque-lumen interface, and is a dimensionless parameter. Blood flow velocity was obtained through computational fluid dynamics simulation based on the realistic lumen geometry, expressed in centimeters per second (cm / s).

[0034] Based on the extracted geometric and physical characteristic data, the comprehensive burden index of coronary artery plaques is calculated using the following formula:

[0035] ;

[0036] Among them, CLI is the overall plaque load index; Vb is the plaque volume, measured in cubic millimeters (mm³), representing the physical space occupied by the plaque; a higher value indicates a larger plaque volume; Db is the plaque mean density, measured in Hounsfield units (HU), reflecting the composition of the plaque; a higher value indicates more calcified plaque; Rs is the plaque surface roughness, dimensionless, reflecting the irregularity of the plaque surface; a higher value indicates a rougher plaque surface; Sn is the luminal stenosis, measured as a percentage (%), representing the degree of obstruction in the luminal cross-sectional area; a higher value indicates more severe stenosis; Vf is the average blood flow velocity at the stenosis, measured in centimeters per second (cm / s), reflecting the hemodynamic state; a higher value indicates a faster blood flow velocity; Vref vol The reference lumen volume is set to 100 mm³, used to normalize the plaque volume; Dref is the reference density, set to 100 HU, used to normalize the plaque density; Vref flow The reference flow rate is set to 50 cm / s, which is used to normalize the blood flow velocity.

[0037] The calculation principle of this formula is based on a biophysical model of the impact of plaque load on hemodynamics. By normalizing the ratio of plaque volume, density, surface irregularity, and the stenosis effect caused by the plaque to changes in blood flow velocity, a dimensionless comprehensive load index is generated. The higher the index value, the greater the potential impact of the plaque on coronary blood flow and the higher the clinical risk.

[0038] Based on the calculated comprehensive burden index and a preset burden risk threshold, a graded assessment is performed to generate plaque burden risk levels, which are divided into three levels: low risk, medium risk, and high risk. Low risk corresponds to a comprehensive burden index less than 0.1, indicating a mild plaque burden with minimal impact on blood flow; medium risk corresponds to a comprehensive burden index between 0.1 and 0.3, indicating a moderate plaque burden that has a clear impact on blood flow; high risk corresponds to a comprehensive burden index greater than 0.3, indicating a severe plaque burden that poses a significant threat to coronary artery function and may require aggressive clinical intervention.

[0039] Based on the grading assessment results, a detailed quantitative assessment report of plaque burden is generated. The report includes the comprehensive burden index value, the corresponding risk level, key parameter analysis, and risk-based clinical intervention recommendations. This module provides objective and quantitative evidence for the clinical diagnosis and treatment decisions of coronary artery plaques and stenosis by integrating multiple parameters into a single, dimensionless comprehensive burden index and performing risk grading, which significantly improves the accuracy and repeatability of diagnosis.

[0040] Preferably, the photon CT pure calcification sequence analysis system for coronary artery plaques and stenosis also includes a hemodynamic optimization analysis module for performing the following steps:

[0041] Hemodynamic data of the coronary arteries were extracted from artifact-free, real-world views of the lumen, including lumen diameter, blood flow velocity, wall shear stress, and turbulence index. The lumen diameter was obtained through geometric reconstruction of the real lumen view, accurate to 0.01 mm. Blood flow velocity was calculated by simulating the hemodynamic distribution based on feature data from pure calcified CT sequence images, measured in centimeters per second. Wall shear stress was calculated using the frictional force between the blood flow velocity and the lumen wall, measured in Pascals. The turbulence index was determined by analyzing the spatial rate of change and directionality of the blood flow velocity, and is a dimensionless parameter.

[0042] Using the extracted hemodynamic data, a coronary artery hemodynamic model was constructed, and a three-dimensional simulation of blood flow was performed using computational fluid dynamics methods to generate blood flow velocity distribution maps and wall shear stress distribution maps.

[0043] By analyzing blood flow velocity distribution maps and wall shear stress distribution maps, we can identify potential high-shear stress and low-shear stress regions within the coronary arteries and determine whether they are related to plaque formation or worsening stenosis.

[0044] By combining luminal diameter and blood flow turbulence index, the impact of hemodynamic abnormalities on coronary plaque stability is assessed, and the areas of hemodynamic abnormalities are identified.

[0045] Quantitative analysis of hemodynamically abnormal areas is performed to calculate the area ratio and impact of abnormal areas, generating a hemodynamic analysis report. This report includes blood flow velocity distribution, wall shear stress distribution, turbulence index distribution, and a detailed description of the abnormal areas, and provides a correlation analysis between the abnormal areas and plaques and stenosis.

[0046] Compared with the prior art, the beneficial effects of the present invention are:

[0047] This invention converts incident X-ray photons into electrical signals using a semiconductor detector. Utilizing the differences in X-ray absorption characteristics between calcification, blood, and soft tissue at different energies, and leveraging the energy resolution of photon-counting CT, it directly separates and quantifies calcified tissue from the raw data level, acquiring pure calcification sequence image data from photon-counting CT. This pure calcification sequence image data is processed to construct a coronary artery calcification artifact correction model. Based on this model, CCTA images are analyzed and identified, and coronary artery calcification artifacts are precisely corrected to determine a true lumen view free of artifact interference. Based on this true lumen view, coronary artery plaques and stenosis are analyzed and diagnosed, accurately assessing the condition and generating an analysis and assessment report. This allows for precise analysis of coronary artery plaques and stenosis using pure calcification sequences from photon-counting CT, improving the accuracy of coronary artery plaque and stenosis analysis and diagnosis. Attached Figure Description

[0048] Figure 1 This is a block diagram of the photonic CT pure calcification sequence analysis system for coronary artery plaques and stenosis according to the present invention;

[0049] Figure 2 This is a flowchart of the system for analyzing coronary artery plaques and stenosis using pure calcification sequences from photon CT according to the present invention. Detailed Implementation

[0050] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0051] To address the current limitations in accurately analyzing coronary artery plaques and stenosis using pure calcification sequences from photon CT, which leads to low accuracy in the analysis and diagnosis of coronary artery plaques and stenosis, please refer to [link to relevant documentation]. Figures 1-2 This embodiment provides the following technical solution:

[0052] A system for analyzing coronary artery plaques and stenosis using pure calcification sequences from photon CT, including:

[0053] The data acquisition sequence generation module is used to perform coronary artery scanning based on photon counting CT and acquire pure calcification sequence image data from photon CT.

[0054] In this embodiment, coronary artery scanning is performed based on photon counting CT to acquire pure calcification sequence image data from photon CT, and the following operations are performed:

[0055] Based on semiconductor detectors, incident X-ray photons are converted into electrical signals. By utilizing the differences in X-ray absorption characteristics of calcification, blood, and soft tissue at different energies, and leveraging the energy resolution capability of photon counting CT, calcified tissue is directly separated and quantified from the raw data level. Signals that conform to the energy characteristics of calcification are found and retained, while signals that do not conform to the energy characteristics of calcification are suppressed, generating a pure calcification sequence. This allows for the acquisition of pure calcification sequence image data from photon CT, displaying only calcification information without interference from lumens and soft tissues, and providing precise three-dimensional spatial distribution and density information of calcified plaques.

[0056] It should be noted that a semiconductor detector is a radiation detector that uses semiconductor materials as the detection medium. The semiconductor material used in this semiconductor detector is cadmium telluride. The basic principle of a semiconductor detector is that charged particles generate electron-hole pairs within the sensitive volume of the semiconductor detector. These electron-hole pairs drift under the influence of an external electric field and output a signal. The pulse height of the electrical signal generated by each incident X-ray photon is proportional to the photon energy. By setting multiple energy thresholds, photons in different energy ranges can be counted and classified. Through a basic material decomposition algorithm, the detected multi-energy signal is decomposed into a combination of basic materials such as calcification and iodine, thereby generating an image sequence containing only calcification signals. This completely eliminates the influence of blood vessels and surrounding soft tissues, thus obtaining pure calcification sequence image data for photon CT.

[0057] The calcification artifact precision correction module is used to process pure calcification sequence image data from photon CT and construct a coronary artery calcification artifact correction model. It performs precise correction of coronary artery calcification artifacts on CCTA images to determine the true lumen view without artifact interference.

[0058] It should be noted that severe calcifications on traditional CT scans can cause artifacts, also known as partial volume effects, making the calcifications appear larger than they actually are and obscuring adjacent lumens, leading to an overestimation of stenosis. Therefore, registering and fusing clean calcification sequences with traditional CCTA images can accurately correct coronary artery calcification artifacts in CCTA images, determine a true lumen view free from artifact interference, and facilitate more accurate analysis and assessment of coronary artery plaques and stenosis.

[0059] In this embodiment, the pure calcification sequence image data from photon CT is processed by performing the following operations:

[0060] The pure calcification sequence image data of photon CT is cleaned to remove noise, and missing and outlier values ​​in the pure calcification sequence image data of photon CT are identified and evaluated.

[0061] Among them, it is necessary to determine whether missing and outlier values ​​in pure calcification sequence image data of photon CT are valuable for the analysis of coronary artery plaques and stenosis, and to process the missing and outlier values ​​in pure calcification sequence image data of photon CT according to the determination.

[0062] When missing and outlier values ​​in pure calcification sequence image data of photon CT are valuable for the analysis of coronary artery plaques and stenosis, the median is used to fill the missing values ​​in the pure calcification sequence image data of photon CT, and the mean is used to replace the outlier values ​​in the pure calcification sequence image data of photon CT.

[0063] When missing and outlier values ​​in the pure calcification sequence image data of photon CT are of no value for the analysis of coronary artery plaques and stenosis, then the missing and outlier values ​​in the pure calcification sequence image data of photon CT are removed.

[0064] It should be noted that by cleaning the pure calcification sequence image data of photon CT, removing noise from the pure calcification sequence image data, and processing the missing and outlier values ​​in the identified pure calcification sequence image data, the data quality of the pure calcification sequence image data of photon CT can be improved, which facilitates better subsequent analysis of coronary artery plaques and stenosis.

[0065] In this embodiment, the pure calcification sequence image data from photon CT is processed, and the following operations are performed:

[0066] Enhancement processing is performed on pure calcification sequence image data from photon CT to emphasize the overall or local characteristics of the image, make the originally unclear image clearer, amplify the differences between object features in the image, and preserve the sharp edges of coronary artery calcification plaques;

[0067] Feature extraction was performed on pure calcification sequence image data from photon CT. Features related to coronary artery plaque and stenosis analysis were extracted from the pure calcification sequence image data from photon CT, and the feature data of pure calcification sequence image from photon CT were determined to facilitate better subsequent analysis of coronary artery plaque and stenosis.

[0068] In this embodiment, precise correction of coronary artery calcification artifacts is performed on CCTA images by performing the following operations:

[0069] A coronary artery calcification artifact correction model was constructed based on the feature data of pure calcification sequence images from photon CT. CCTA images were then input into the coronary artery calcification artifact correction model. The CCTA images were analyzed and identified according to the coronary artery calcification artifact correction model, and the coronary artery calcification artifacts were accurately corrected to determine the true lumen view without artifact interference, clearly displaying the real vascular lumen.

[0070] In this embodiment, a coronary artery calcification artifact correction model is constructed, and the following operations are performed:

[0071] The feature data of pure calcification sequence images in photon CT are divided into training set and test set.

[0072] The machine learning model is trained using a training set, enabling it to autonomously learn the coronary artery calcification artifact correction behavior from the training set and accurately correct coronary artery calcification artifacts, thus determining the coronary artery calcification artifact correction model.

[0073] The coronary artery calcification artifact correction model was tested using a test set, and its generalization performance was evaluated to determine the model test evaluation results.

[0074] The coronary artery calcification artifact correction model was optimized based on the model test evaluation results to determine the optimal coronary artery calcification artifact correction model.

[0075] In this embodiment, the generalization performance of the coronary artery calcification artifact correction model is evaluated by performing the following operations:

[0076] The test set was input into the coronary artery calcification artifact correction model. The generalization performance of the coronary artery calcification artifact correction model was evaluated based on accuracy and F1 score, and it was determined whether the coronary artery calcification artifact correction model could achieve the expected effect of accurately correcting coronary artery calcification artifacts.

[0077] When the coronary artery calcification artifact correction model fails to achieve the expected effect of accurately correcting coronary artery calcification artifacts, the parameters of the coronary artery calcification artifact correction model are adjusted and iteratively optimized until the coronary artery calcification artifact correction model can achieve the expected effect of accurately correcting coronary artery calcification artifacts, and the optimal coronary artery calcification artifact correction model is determined.

[0078] The plaque and stenosis analysis and diagnosis module is used to analyze and diagnose coronary artery plaques and stenosis based on real lumen views, accurately assessing the status of coronary artery plaques and stenosis.

[0079] In this embodiment, coronary artery plaques and stenosis are analyzed based on real lumen views without artifact interference to assess the degree of stenosis. Quantitative analysis of non-calcified plaques and low-density plaques is also performed to provide a more comprehensive plaque burden assessment and to determine the analysis and assessment report of coronary artery plaques and stenosis.

[0080] In summary, based on the conversion of incident X-ray photons into electrical signals using semiconductor detectors, and utilizing the differences in X-ray absorption characteristics of calcification, blood, and soft tissue at different energies, photon-counting CT directly separates and quantifies calcified tissue from the raw data level based on its energy resolution capability. This yields pure calcification sequence image data from photon-counting CT. The pure calcification sequence image data is then processed to construct a coronary artery calcification artifact correction model. Based on this model, CCTA images are analyzed and identified, and coronary artery calcification artifacts are precisely corrected to determine a true lumen view free of artifact interference. Based on this true lumen view, coronary artery plaques and stenosis are analyzed and diagnosed, accurately assessing the condition and generating an analysis and assessment report. This method allows for precise analysis of coronary artery plaques and stenosis using pure calcification sequences from photon-counting CT, thus improving the accuracy of coronary artery plaque and stenosis analysis and diagnosis.

[0081] In this embodiment, the system for analyzing coronary artery plaques and stenosis using pure calcification sequences from photon CT is characterized by further including a plaque burden quantitative assessment module, used to perform the following steps:

[0082] Geometric and physical characteristics of coronary artery plaques were extracted from artifact-free, realistic lumen views, including plaque volume, average plaque density, luminal stenosis, plaque surface roughness, and blood flow velocity. Plaque volume was calculated using 3D reconstruction of the plaque region in the realistic lumen view, expressed in millimeters per cubic millimeter (mm³). Average plaque density was determined by calculating the average grayscale value of the plaque region in clean calcified CT images, expressed in Hounsfield units (HU). Lumen stenosis was calculated as a percentage (%) of the normal lumen diameter compared to the minimum diameter at the stenosis. Plaque surface roughness was determined by analyzing the pixel intensity variation rate at the plaque-lumen interface, and is a dimensionless parameter. Blood flow velocity was obtained through computational fluid dynamics simulation based on the realistic lumen geometry, expressed in centimeters per second (cm / s).

[0083] Based on the extracted geometric and physical characteristic data, the comprehensive burden index of coronary artery plaques is calculated using the following formula:

[0084] ;

[0085] Among them, CLI is the overall plaque load index; Vb is the plaque volume, measured in cubic millimeters (mm³), representing the physical space occupied by the plaque; a higher value indicates a larger plaque volume; Db is the plaque mean density, measured in Hounsfield units (HU), reflecting the composition of the plaque; a higher value indicates more calcified plaque; Rs is the plaque surface roughness, dimensionless, reflecting the irregularity of the plaque surface; a higher value indicates a rougher plaque surface; Sn is the luminal stenosis, measured as a percentage (%), representing the degree of obstruction in the luminal cross-sectional area; a higher value indicates more severe stenosis; Vf is the average blood flow velocity at the stenosis, measured in centimeters per second (cm / s), reflecting the hemodynamic state; a higher value indicates a faster blood flow velocity; Vref vol The reference lumen volume is set to 100 mm³, used to normalize the plaque volume; Dref is the reference density, set to 100 HU, used to normalize the plaque density; Vref flow The reference flow rate is set to 50 cm / s, which is used to normalize the blood flow velocity.

[0086] The calculation principle of this formula is based on a biophysical model of the impact of plaque load on hemodynamics. By normalizing the ratio of plaque volume, density, surface irregularity, and the stenosis effect caused by the plaque to changes in blood flow velocity, a dimensionless comprehensive load index is generated. The higher the index value, the greater the potential impact of the plaque on coronary blood flow and the higher the clinical risk.

[0087] Based on the calculated comprehensive burden index and a preset burden risk threshold, a graded assessment is performed to generate plaque burden risk levels, which are divided into three levels: low risk, medium risk, and high risk. Low risk corresponds to a comprehensive burden index less than 0.1, indicating a mild plaque burden with minimal impact on blood flow; medium risk corresponds to a comprehensive burden index between 0.1 and 0.3, indicating a moderate plaque burden that has a clear impact on blood flow; high risk corresponds to a comprehensive burden index greater than 0.3, indicating a severe plaque burden that poses a significant threat to coronary artery function and may require aggressive clinical intervention.

[0088] Based on the grading assessment results, a detailed quantitative assessment report of plaque burden is generated. The report includes the comprehensive burden index value, the corresponding risk level, key parameter analysis, and risk-based clinical intervention recommendations. This module provides objective and quantitative evidence for the clinical diagnosis and treatment decisions of coronary artery plaques and stenosis by integrating multiple parameters into a single, dimensionless comprehensive burden index and performing risk grading, which significantly improves the accuracy and repeatability of diagnosis.

[0089] The plaque load quantitative assessment module, based on a real lumen view free of artifact interference, achieves quantitative assessment of plaque load through multi-step data extraction and calculation. First, the plaque volume is obtained through 3D reconstruction calculation. Specifically, a region growing segmentation algorithm is used to automatically delineate the plaque boundaries in the real lumen view. The total volume occupied by the plaque is calculated based on voxel counting (voxel size 0.3mm × 0.3mm × 0.3mm) and a spatial calibration model, expressed in cubic millimeters (mm³). The average plaque density is determined by analyzing the pixel grayscale values ​​of segmented plaque regions in pure calcification sequence images from photon CT. Specifically, the Hounsfield unit (HU) arithmetic mean of all plaque pixels is calculated using the grayscale histogram statistical method to exclude noise interference outside the calcification region. Lumen stenosis was calculated by comparing the normal lumen diameter with the minimum diameter at the stenosis point. A vascular path planning algorithm based on centerline extraction was used to automatically locate the normal segment (≥5mm from the plaque) and the stenotic segment on the actual lumen view. Their diameters were measured and calculated using the formula (1 - minimum diameter at the stenosis point / normal segment diameter) × 100%, expressed as a percentage (%). Plaque surface roughness was determined by analyzing the pixel intensity change rate at the plaque-lumen interface. The Sobel operator was used to calculate the gradient amplitude of the interface region, and then normalized (gradient amplitude / maximum possible gradient value) to obtain a dimensionless parameter. Blood flow velocity was obtained through computational fluid dynamics simulation. A three-dimensional blood flow field was established based on the actual lumen geometry model. A Newtonian fluid model with a blood density of 1060 kg / m³ and a viscosity of 0.0035 Pa•s was set. Patient-specific blood pressure boundary conditions (systolic blood pressure 120 mmHg / diastolic blood pressure 80 mmHg) were applied to solve the Navier-Stokes equations, and the average flow velocity of the stenotic segment was extracted, expressed as centimeters per second (cm / s).

[0090] Based on the above parameters, the comprehensive load index formula is adopted. Calculations are performed, including reference values. , , For parameter normalization, this formula generates a dimensionless index by quantifying the interaction between plaque physical properties and hemodynamics; the higher the value, the greater the overall impact of the plaque on coronary blood flow. Finally, risk stratification is performed based on CLI values ​​(low risk <0.1, 0.1≤medium risk≤0.3, high risk>0.3), and an assessment report including numerical analysis, risk level, and clinical recommendations is generated.

[0091] To better illustrate this embodiment, the following specific implementation examples are provided:

[0092] In clinical application, a real lumen view of a 65-year-old male patient with coronary artery disease was analyzed. The view data was loaded using a medical image processing platform (such as 3DSlicer), and plaque 3D reconstruction was completed through threshold segmentation (HU>130) and morphological closure operation, calculating the plaque volume to be 85 mm³. The plaque region ROI (128×128 pixels) was extracted from the pure calcified sequence, with a statistically average density of 285 HU. The normal lumen diameter (3.2 mm) and the minimum diameter at the stenosis (1.9 mm) were measured along the vessel centerline, calculating a stenosis degree of 40.6%. Pixel gradient analysis (kernel size 3×3) was performed on the plaque-lumen interface, calculating a surface roughness of 0.18. A CFD simulation (ANSYS Fluent software) with a transient solver (time step 0.01 s) was used to simulate an average blood flow velocity of 42 cm / s at the stenosis. Substituting these values ​​into the formula... This is classified as a high-risk case. During the report generation phase, the system automatically integrates parameters (plaque volume accounting for 85% of the normal lumen, high calcification density, slightly irregular surface, and moderately accelerated blood flow), recommends active clinical interventions such as drug therapy or revascularization, and marks the location of the high-risk plaque (mid-segment of the left anterior descending artery) using a color-coded 3D model, providing intuitive evidence for clinical decision-making.

[0093] In this embodiment, the photon CT pure calcification sequence analysis system for coronary artery plaques and stenosis further includes a hemodynamic optimization analysis module for performing the following steps:

[0094] Hemodynamic data of the coronary arteries were extracted from artifact-free, real-world views of the lumen, including lumen diameter, blood flow velocity, wall shear stress, and turbulence index. The lumen diameter was obtained through geometric reconstruction of the real lumen view, accurate to 0.01 mm. Blood flow velocity was calculated by simulating the hemodynamic distribution based on feature data from pure calcified CT sequence images, measured in centimeters per second. Wall shear stress was calculated using the frictional force between the blood flow velocity and the lumen wall, measured in Pascals. The turbulence index was determined by analyzing the spatial rate of change and directionality of the blood flow velocity, and is a dimensionless parameter.

[0095] Using the extracted hemodynamic data, a coronary artery hemodynamic model was constructed, and a three-dimensional simulation of blood flow was performed using computational fluid dynamics methods to generate blood flow velocity distribution maps and wall shear stress distribution maps.

[0096] By analyzing blood flow velocity distribution maps and wall shear stress distribution maps, we can identify potential high-shear stress and low-shear stress regions within the coronary arteries and determine whether they are related to plaque formation or worsening stenosis.

[0097] By combining luminal diameter and blood flow turbulence index, the impact of hemodynamic abnormalities on coronary plaque stability is assessed, and the areas of hemodynamic abnormalities are identified.

[0098] Quantitative analysis of hemodynamically abnormal areas is performed to calculate the area ratio and impact of abnormal areas, generating a hemodynamic analysis report. This report includes blood flow velocity distribution, wall shear stress distribution, turbulence index distribution, and a detailed description of the abnormal areas, and provides a correlation analysis between the abnormal areas and plaques and stenosis.

[0099] The hemodynamic optimization analysis module achieves precise localization of blood flow abnormalities by extracting multidimensional hemodynamic parameters from the actual lumen view and establishing a physical model. The lumen inner diameter is obtained through geometric reconstruction measurement. A 3D surface reconstruction technique based on the MarchingCubes algorithm is used to convert the actual lumen view into a mesh model (resolution 0.01 mm). The inner diameter is automatically measured on a cross-section perpendicular to the vessel centerline using the minimum circumscribed circle algorithm, accurate to 0.01 mm. Blood flow velocity is calculated through hemodynamic simulation based on the feature data of pure calcification sequence images from photon CT. Specifically, the actual lumen geometric model is imported into finite element analysis software (such as COMSOL), a non-Newtonian Carreau model is set to describe the blood rheological properties, and a patient-specific cardiac cycle flow curve (calibrated by Doppler ultrasound) is applied as the inlet boundary condition. After solving the transient flow equation, the time-averaged velocity is extracted, in centimeters per second. Wall shear stress is calculated through the blood flow velocity and the frictional force on the lumen wall. The gradient calculation function in the CFD post-processing module is used to calculate the stress within the wall boundary layer based on the velocity gradient (du / dy) and the blood dynamic viscosity (μ). The formula calculates the transient value, where μ is the blood dynamic viscosity, taken as 0.0035 Pa·s. The time-averaged value of the transient value is then used to obtain the time-averaged wall shear stress, in Pascals (Pa). The blood flow turbulence index is determined by analyzing the spatial rate of change and directionality of blood flow velocity. The normalized value of the velocity fluctuation energy (turbulence intensity) is calculated using the Reynolds stress decomposition method. Specifically, an evaluation region is selected in the simulated flow field, and the ratio of the standard deviation of the velocity vector to the average velocity is calculated to obtain the dimensionless parameter. When constructing the coronary artery hemodynamic model using the above data, the finite volume method is used to discretize the computational domain (grid number ≥ 100,000), and the SIMPLE algorithm is used to solve the pressure-velocity coupled field, generating a blood flow velocity distribution map (color-mapped velocity range 0-100 cm / s) and a wall shear stress distribution map (color-mapped stress range 0-10 Pa). When identifying abnormal regions through distribution map analysis, a high shear stress threshold of >2.5 Pa (correlated with endothelial cell damage) and a low shear stress threshold of <0.4 Pa (correlated with lipid deposition) were set. The correlation between hemodynamic abnormalities and plaque stability was assessed by combining the local coefficient of variation of the luminal diameter (>15%) and the turbulence index (>0.2). In the final quantitative analysis stage, a region growing algorithm was used to segment abnormal regions, calculate their percentage of the total vascular surface area, and assess the degree of influence by combining peak shear stress and maximum turbulence intensity, generating a report containing multidimensional parameter correlation analysis.

[0100] To better illustrate this embodiment, the following specific implementation examples are provided:

[0101] In a specific case analysis, a real lumen view of a patient with right coronary artery stenosis was processed. After obtaining a vascular model through 3D reconstruction, the lumen diameter at the distal end of the stenosis was measured to be 1.58 mm (normal segment 2.84 mm). CFD simulation was performed with periodic boundary conditions (heart rate 75 bpm, period 0.8 s). The simulation results showed that the jet velocity at the stenosis reached 68 cm / s, while the velocity in the downstream vortex region decreased to 12 cm / s. Based on the blood dynamic viscosity μ = 0.0035 Pa·s, the wall shear stress was calculated. The analysis showed that the stress at the stenosis throat reached 4.2 Pa (high shear zone), while the stress in the dilated region was only 0.3 Pa (low shear zone). The turbulence index was calculated as 0.36 by the ratio of the velocity fluctuation (standard deviation ± 9 cm / s) in the downstream vortex region to the average velocity (25 cm / s). The analysis determined that the high shear zone was associated with potential plaque rupture (area share 8%), while the low shear zone was associated with the risk of new plaque formation (area share 22%). When the report is generated, the system integrates parameters to generate velocity streamline diagrams, stress cloud diagrams, and turbulence isosurface diagrams, marks the spatial coordinates of the abnormal area (47-53 mm from the coronary artery ostium), and calculates through a logistic regression model that the abnormality increases the risk of plaque instability by 3.2 times, recommending priority treatment of this segment and optimization of the antiplatelet therapy regimen.

[0102] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0103] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A system for analyzing coronary artery plaques and stenosis using pure calcification sequences from photon CT, characterized in that, include: The data acquisition sequence generation module is used to perform coronary artery scanning based on photon counting CT and acquire pure calcification sequence image data from photon CT. The calcification artifact precision correction module is used to process pure calcification sequence image data from photon CT and construct a coronary artery calcification artifact correction model. It performs precise correction of coronary artery calcification artifacts on CCTA images to determine the true lumen view without artifact interference. The plaque and stenosis analysis and diagnosis module is used to analyze and diagnose coronary artery plaques and stenosis based on real lumen views, and to accurately assess the status of coronary artery plaques and stenosis. It also includes a plaque load quantitative assessment module for performing the following steps: Geometric and physical characteristics of coronary artery plaques were extracted from artifact-free, realistic lumen views, including plaque volume, average plaque density, luminal stenosis, plaque surface roughness, and blood flow velocity. Plaque volume was obtained through 3D reconstruction of the plaque region in the realistic lumen view; average plaque density was determined by calculating the average grayscale value of the plaque region in a clean calcified CT sequence; luminal stenosis was calculated by comparing the percentage of the normal lumen diameter to the minimum diameter at the stenosis; plaque surface roughness was determined by analyzing the pixel intensity change rate at the plaque-lumen interface; and blood flow velocity was obtained through computational fluid dynamics simulations based on the realistic lumen geometry. Based on the extracted geometric and physical characteristic data, the comprehensive burden index of coronary artery plaques is calculated using the following formula: ； Among them, CLI is the overall plaque load index; Vb is the plaque volume in cubic millimeters, representing the physical space occupied by the plaque; a higher value indicates a larger plaque volume; Db is the plaque mean density in Hounsfield units, reflecting the composition of the plaque; a higher value indicates more calcified plaque; Rs is the plaque surface roughness, dimensionless, reflecting the irregularity of the plaque surface; a higher value indicates a rougher plaque surface; Sn is the luminal stenosis, in percentage, representing the degree of obstruction of the luminal cross-sectional area; a higher value indicates more severe stenosis; Vf is the average blood flow velocity at the stenosis site in centimeters per second, reflecting the hemodynamic state; a higher value indicates a faster blood flow velocity; Vref vol The reference lumen volume is set to 100 mm³, used to normalize the plaque volume; Dref is the reference density, set to 100 HU, used to normalize the plaque density; Vref flow The reference flow rate is set to 50 cm / s, which is used to normalize the blood flow velocity. Based on the calculated comprehensive load index, and combined with the preset load risk threshold, a graded assessment is conducted to generate patch load risk levels, which are divided into three levels: low risk, medium risk, and high risk. Based on the grading assessment results, a detailed quantitative assessment report of plaque burden is generated. The report includes the comprehensive burden index value, the corresponding risk level, key parameter analysis, and risk-based clinical intervention recommendations.

2. The system for analyzing coronary plaque and stenosis according to claim 1, wherein the photon CT pure calcification sequence is a dual-energy photon CT pure calcification sequence. Coronary artery scanning was performed using photon counting CT to acquire clean calcification sequence image data. The following operations were then performed: Based on semiconductor detectors, incident X-ray photons are converted into electrical signals. By utilizing the differences in X-ray absorption characteristics of calcification, blood, and soft tissue at different energies, and leveraging the energy resolution capability of photon counting CT, calcified tissue is directly separated and quantified from the raw data level. Signals that conform to the energy characteristics of calcification are found and retained, while signals that do not conform to the energy characteristics of calcification are suppressed, generating a pure calcification sequence. This allows the acquisition of pure calcification sequence image data from photon CT, displaying only calcification information without interference from lumens and soft tissues.

3. The system for analyzing coronary plaque and stenosis according to claim 2, wherein the photon CT pure calcification sequence is a dual-energy photon CT pure calcification sequence. The following operations are performed on the purified calcification sequence image data from photon CT: The pure calcification sequence image data of photon CT is cleaned to remove noise, and missing and outlier values ​​in the pure calcification sequence image data of photon CT are identified and evaluated. Among them, it is necessary to determine whether missing values ​​and outliers in the pure calcification sequence image data of photon CT are valuable for the analysis of coronary artery plaques and stenosis, and to process the missing values ​​and outliers in the pure calcification sequence image data of photon CT according to the determination. When missing and outlier values ​​in pure calcification sequence image data of photon CT are valuable for the analysis of coronary artery plaques and stenosis, the median is used to fill the missing values ​​in the pure calcification sequence image data of photon CT, and the mean is used to replace the outlier values ​​in the pure calcification sequence image data of photon CT. When missing and outlier values ​​in the pure calcification sequence image data of photon CT are of no value for the analysis of coronary plaques and stenosis, then the missing and outlier values ​​in the pure calcification sequence image data of photon CT are removed.

4. The system for analyzing coronary plaque and stenosis according to claim 3, wherein the photon CT pure calcification sequence is a dual-energy photon CT pure calcification sequence. The pure calcification sequence image data from photon CT is processed, and the following operations are performed: Enhancement processing is performed on pure calcification sequence image data from photon CT to emphasize the overall or local characteristics of the image, make the originally unclear image clearer, amplify the differences between object features in the image, and preserve the sharp edges of coronary artery calcification plaques; Feature extraction was performed on pure calcification sequence image data from photon CT, and features related to coronary artery plaque and stenosis analysis were extracted from the pure calcification sequence image data from photon CT to determine the feature data of pure calcification sequence images from photon CT.

5. The system for analyzing coronary plaque and stenosis according to claim 4, wherein the photon CT pure calcification sequence is a dual-energy photon CT pure calcification sequence. To accurately correct coronary artery calcification artifacts on CCTA images, perform the following operations: A coronary artery calcification artifact correction model was constructed based on the feature data of pure calcification sequence images from photon CT. CCTA images were then input into the coronary artery calcification artifact correction model. The CCTA images were analyzed and identified according to the coronary artery calcification artifact correction model, and the coronary artery calcification artifacts were accurately corrected to determine the true lumen view without artifact interference, clearly displaying the real vascular lumen.

6. The system for analyzing coronary plaque and stenosis according to claim 5, wherein the photon CT pure calcification sequence is a dual-energy photon CT pure calcification sequence. To construct a coronary artery calcification artifact correction model, perform the following operations: The feature data of pure calcification sequence images in photon CT are divided into training set and test set. The machine learning model is trained using a training set, enabling it to autonomously learn the coronary artery calcification artifact correction behavior from the training set and accurately correct coronary artery calcification artifacts, thus determining the coronary artery calcification artifact correction model. The coronary artery calcification artifact correction model was tested using a test set, and its generalization performance was evaluated to determine the model test evaluation results. The coronary artery calcification artifact correction model was optimized based on the model test evaluation results to determine the optimal coronary artery calcification artifact correction model.

7. The system for analyzing coronary plaque and stenosis according to claim 6, wherein the photon CT pure calcification sequence is a dual-energy photon CT pure calcification sequence. To evaluate the generalization performance of the coronary artery calcification artifact correction model, perform the following operations: The test set was input into the coronary artery calcification artifact correction model. The generalization performance of the coronary artery calcification artifact correction model was evaluated based on accuracy and F1 score, and it was determined whether the coronary artery calcification artifact correction model could achieve the expected effect of accurately correcting coronary artery calcification artifacts. When the coronary artery calcification artifact correction model fails to achieve the expected effect of accurately correcting coronary artery calcification artifacts, the parameters of the coronary artery calcification artifact correction model are adjusted and iteratively optimized until the coronary artery calcification artifact correction model can achieve the expected effect of accurately correcting coronary artery calcification artifacts, and the optimal coronary artery calcification artifact correction model is determined.

8. The system for analyzing coronary artery plaques and stenosis using pure photon CT calcification sequences according to claim 7, characterized in that, Analysis of coronary artery plaques and stenosis is performed based on real lumen views free from artifacts, assessing the degree of stenosis and quantitatively analyzing non-calcified plaques and low-density plaques to provide a more comprehensive assessment of plaque burden and determine the analytical assessment report of coronary artery plaques and stenosis.

9. The system for analyzing coronary plaque and stenosis according to claim 8, wherein, It also includes a hemodynamic optimization analysis module, used to perform the following steps: Hemodynamic data of the coronary arteries were extracted from artifact-free, real-world views of the lumen, including luminal diameter, blood flow velocity, wall shear stress, and turbulence index. The luminal diameter was obtained through geometric reconstruction of the real-world lumen view. Blood flow velocity was calculated by simulating the hemodynamic distribution based on feature data from pure calcified CT sequence images. Wall shear stress was calculated using the frictional force between the blood flow velocity and the lumen wall. The turbulence index was determined by analyzing the spatial rate of change and directionality of the blood flow velocity. Using the extracted hemodynamic data, a coronary artery hemodynamic model was constructed, and a three-dimensional simulation of blood flow was performed using computational fluid dynamics methods to generate blood flow velocity distribution maps and wall shear stress distribution maps. By analyzing blood flow velocity distribution maps and wall shear stress distribution maps, we can identify potential high-shear stress and low-shear stress regions within the coronary arteries and determine whether they are related to plaque formation or worsening stenosis. By combining luminal diameter and blood flow turbulence index, the impact of hemodynamic abnormalities on coronary plaque stability is assessed, and the areas of hemodynamic abnormalities are identified. Quantitative analysis of hemodynamically abnormal areas is performed to calculate the area ratio and impact of abnormal areas, generating a hemodynamic analysis report. This report includes blood flow velocity distribution, wall shear stress distribution, turbulence index distribution, and a detailed description of the abnormal areas, and provides a correlation analysis between the abnormal areas and plaques and stenosis.

Images