Method and system for quantitative microvascular dysfunction in sequences of angiographic images

Abstract

A method and system are provided for characterizing microvascular tissue supplied with blood through a coronary artery under examination, comprising acquiring a sequence of x-ray angiographic images of the coronary artery under examination acquired while a contrast agent flows into and through the coronary artery under examination, the angiographic image sequence being used to determine a volumetric flow rate of blood flow through the coronary artery under examination, and the volumetric flow rate being used to determine an index representative of the microvascular tissue supplied with blood through the coronary artery under examination.

Description

[Technical Field]

[0001] The present disclosure relates to methods and systems for capturing and analyzing angiographic images for the assessment of coronary artery disease. [Background technology]

[0002] Coronary artery disease (CAD) is one of the leading causes of death and serious illnesses in Western countries. Patients with CAD experience angina pectoris, the most common symptom of CAD, affecting approximately 112 million people worldwide. The 2019 ECS Guidelines (Non-Patent Document 1) provide guidance on the diagnosis and management of patients with chronic CAD. However, the majority (approximately 50%) of patients with angina symptoms who undergo coronary angiography have evidence of ischemia but no obstructive coronary artery disease. Many of these patients have normal or near-normal coronary angiography results, which makes diagnosis difficult for clinicians. This condition is called dysfunction of the heart's small blood vessels (microvascular dysfunction), specifically coronary microvascular dysfunction, which is related to the coronary circulation. Coronary microvascular dysfunction primarily affects the heart's arterioles and capillaries. These vessels regulate blood flow within the myocardium and ensure the heart receives sufficient oxygen and nutrients. Coronary microvascular dysfunction is often considered a significant problem in patients with angina due to obstructive coronary artery disease (ANOCA). Coronary microvascular dysfunction is more prevalent in women than in men (Non-Patent Document 2). Coronary microvascular dysfunction is an active area of ​​research in cardiology, and our understanding of its causes, diagnosis, and management is evolving. It is important for individuals with symptoms suggestive of heart disease to seek medical advice, and for healthcare professionals to consider coronary microvascular dysfunction as a potential cause, especially when conventional testing does not provide a clear diagnosis.

[0003] Patients with CAD disease are primarily treated with percutaneous coronary intervention (PCI). PCI is a non-surgical procedure performed in a catheterization laboratory using X-ray angiography. It involves the insertion of small structures called stents and the use of catheters (thin, flexible tubes) to widen heart vessels narrowed by plaque buildup (a condition called atherosclerosis). The symptoms of angina pectoris are thought to be caused by the obstruction of oxygen delivery to the heart muscle due to the narrowing of the coronary arteries. As mentioned above, many patients with symptoms of myocardial ischemia (angina pectoris) do not have obstructive CAD, and coronary microvascular dysfunction is often unrecognized and inadequately treated (Non-Patent Document 3).

[0004] Currently, in the catheterization laboratory, interventional cardiologists can assess microvascular dysfunction using bolus thermodilution or continuous thermodilution (NPL 4). Both techniques require inserting a wire equipped with pressure and temperature sensors into a coronary artery and inducing hyperemia. The index of microvascular resistance (IMR) is an established method for assessing microvascular disease and is based on invasive bolus thermodilution measurements. IMR is a dimensionless index and is calculated by multiplying the distal coronary artery pressure at peak hyperemia by the mean transit time. Both measurements are performed by inserting a wire into a coronary artery, injecting cold saline, measuring the pressure and temperature differences, and extracting the mean transit time. Another invasive approach is based on a Doppler pressure wire and measures hyperemic microvascular resistance (HMR) during hyperemia. HMR is defined as the distal pressure divided by the simultaneously measured flow velocity during hyperemia (NPL 5). However, these techniques are not widely used in routine clinical practice due to technical challenges, procedure costs, increased procedure time, and the intolerance of some patients to hyperemia. [Prior art documents] [Patent documents]

[0005] [Patent Document 1] U.S. Patent No. 11,707,242 "Method and System for Dynamic Coronary Artery Roadmapping" [Patent Document 2] U.S. Patent Application No. 15 / 971,275, "Method and Apparatus for Determining Blood Flow Velocity in X-Ray Angiographic Images" [Patent Document 3] U.S. Patent Application No. 16 / 739,718, "Method and System for Dynamic Coronary Artery Roadmapping" [Patent Document 4] U.S. Patent Application No. 15 / 551,162, "Method and Apparatus for Quantitative Flow Analysis" [Patent Document 5] U.S. Patent Application No. 16 / 438,955, "Quantitative Hemodynamic Flow Analysis Method and Apparatus" [Patent Document 6] U.S. Patent Application No. 17 / 208,373, "Method and System for Aligning Intra-Object and Extra-Object Data" [Non-patent literature]

[0006] [Non-Patent Document 1] Knuuti, Juhani et al., "2019 ESC Guidelines for the Diagnosis and Management of Chronic Coronary Syndromes," European Heart Journal vol. 41,3 (2020): 407-477 [Non-patent document 2] Merz, C.N. et al., "The Women's Ischemia Syndrome Evaluation (WISE) Study: Protocol Design, Methodology, and Feasibility Report," Journal of the American College of Cardiology vol. 33,6 (1999): 1453-61 [Non-patent document 3] Bairey Merz, C Noel et al., "Ischemia and Non-Obstructive Coronary Artery Disease (INOCA): Development of Evidence-Based Treatments and Research Agenda for the Next Decade," Circulation vol. 135,11 (2017): 1075-1092 [Non-patent document 4] Candreva, Alessandro et al., "Fundamentals of Coronary Artery Thrombosis," JACC. Cardiovascular Interventions vol. 14,6 (2021): 595-605 [Non-Patent Document 5] Meuwissen, M. et al., "The role of variations in microvascular resistance on fractional flow reserve and coronary fractional flow velocity reserve in intermediate coronary artery disease," Circulation vol. 103,2 (2001): 184-7 [Non-patent document 6] Zhang, Yimin et al., "Automated coronary blood flow calculation: Validation of quantitative flow ratios from coronary angiography," The International Journal of Cardiovascular Imaging, vol. 35,4 (2019): 587-595. [Non-Patent Document 7] Gronenschild et al., "CAAS. II: A second-generation system for offline and online quantitative coronary angiography," Catheterization and Cardiovascular Diagnosis, vol. 33, 1 (1994): 61-75 [Non-patent document 8] Serife Kaba et al., "Application of Deep Learning for Coronary Artery Segmentation and Classification," Diagnostics 2023, 13, 2274 [Non-Patent Document 9] Qin et al., "High-Precision Binary Image Segmentation," Computer Vision - ECCV 2022 (2022): 38-56 [Non-Patent Document 10] Frangi et al., "Multiscale Vessel Enhancement Filtering," Medical Image Computing and Computer-Assisted Intervention - MICCAI'98 (1998): 130-137 [Non-Patent Document 11] Janssen et al., "A new approach for pathline detection in x-ray angiography: the wavefront propagation algorithm," Int J Cardiovasc Imaging 2002; 18(5):317-324 [Non-Patent Document 12] Zhang et al., "X-ray coronal centerline extraction based on C-UNet and multifactorial reconnection algorithm," Computer Methods and Programs in Biomedicine (2022) 226. 107114 [Non-Patent Document 13] Girasis et al., "Advanced three-dimensional quantitative coronary angiographic assessment of bifurcation lesions: methodology and phantom validation," EuroIntervention 8:1451-1460, 2013 [Non-Patent Document 14] QCA workflow within CAAS Workstation 8.5 (Pie Medical Imaging, the Netherlands) [Non-Patent Document 15] Sen, Sayan et al., "Development and Validation of a New Adenosine-Independent Index of Stenosis Severity from Coronary Wave Intensity Analysis: Results from the ADVISE (Adenosine Vasodilator-Independent Stenosis Evaluation) Study," Journal of the American College of Cardiology vol. 59,15 (2012): 1392-402 [Non-Patent Document 16] Tatineni, S. et al., "Effects of Ionic and Nonionic Radiocontrast Media on Coronary Hyperemia in Patients During Coronary Angiography," American Heart Journal, vol. 123,3 (1992): 621-7 [Non-Patent Document 17] Johnson, Nils P et al., "Does the Instantaneous Wave-Free Ratio Approximate Fractional Flow Reserve?" Journal of the American College of Cardiology vol. 61,13 (2013): 1428-35) [Non-Patent Document 18] Tu, Shengxian et al., "Diagnostic Accuracy of a Rapid Computational Approach to Derive Fractional Flow Reserve from Diagnostic Coronary Angiography: An International Multicenter FAVOR Pilot Study," JACC. Cardiovascular Interventions vol. 9,19 (2016): 2024-2035 [Non-Patent Document 19] Masdjedi et al., "Validation of Fractional Flow Reserve Calculation Software Based on 3D Quantitative Coronary Angiography: The Rapid Assessment of Stenosis Severity (FAST) Study," EuroIntervention: journal of EuroPCR in collaboration with the Working Group on Interventional Cardiology of the European Society of Cardiology 2019 [Non-Patent Document 20] Kirkeeide et al., "Evaluation of Coronary Artery Stenosis by Myocardial Perfusion Imaging During Pharmacological Coronary Artery Dilation. VII. Validation of Flow Reserve as a Single Integrated Functional Index Reflecting All Geometric Aspects of Stenosis Severity," Journal of the American College of Cardiology, vol. 7,1 (1986): 103-13 [Non-Patent Document 21] Gould et al., "Experimental validation of quantitative coronary angiography for determining pressure-flow characteristics of coronary artery stenoses," Circulation vol. 66,5 (1982): 930-7 [Non-Patent Document 22] Thalhammer, Christoph et al., "Noninvasive Central Venous Pressure Measurement by Controlled Compression Ultrasonography in the Forearm," Journal of the American College of Cardiology vol. 50,16 (2007): 1584-9 [Non-Patent Document 23] De Maria et al., "Angiography-derived microcirculatory resistance index as a new, pressure-free tool for assessing coronary microcirculation in ST-segment elevation myocardial infarction," The International Journal of Cardiovascular Imaging, vol. 36,8 (2020): 1395-1406 [Non-Patent Document 24] Scarsini et al., "Angiography-derived index of microcirculatory resistance (IMRangio) as a novel, pressure-free tool for assessing coronary microvascular dysfunction in acute coronary syndromes and stable coronary artery disease," The International Journal of Cardiovascular Imaging, vol. 37, 6 (2021): 1801-1813 [Non-Patent Document 25] Fernandez-Peregrina et al., "Comparison of angiography-derived microcirculatory resistance indexes and invasive assessment indexes in the assessment of coronary microcirculation: a systematic review and meta-analysis," catheterization and cardiovascular interventions: official journal of the Society for Cardiac Angiography & Interventions, vol. 99,7 (2022): 2018-2025 [Non-Patent Document 26] Dodge et al., "Normal Human Coronary Artery Luminal Diameter: Effects of Age, Sex, Anatomical Variations, and Left Ventricular Hypertrophy or Dilatation," Circulation, Vol. 86,1 (1992): 232-46 [Non-Patent Document 27] Choy, Jenny Susana, and Ghassan S Kassab, "Scaling of Myocardial Mass to Coronary Artery Flow and Morphometry," Journal of Applied Physiology (Bethesda, Md.: 1985) vol. 104,5 (2008): 1281-6 [Non-patent document 28] Fearon et al., "A novel index for invasive assessment of coronary microcirculation," Circulation vol. 107,25 (2003): 3129-32 [Non-Patent Document 29] Hao et al., "Vascular Layer Separation in X-Ray Angiography Using Fully Convolutional Networks," Proc. SPIE 10576, Medical Imaging 2018: Image-Guided Procedures, Robotic Interventions, and Modeling [Non-Patent Document 30] Ma et al., "Layer Separation for Vessel Enhancement in Interventional X-Ray Angiography Using Morphological Filtering and Robust PCA," Workshop on Augmented Environments for Computer-Assisted Interventions 2017, Springer. pp. 104-113 [Non-Patent Document 31] De Bruyne et al., "Intracoronary and intravenous adenosine 5'-triphosphate, adenosine, papaverine, and contrast media to assess fractional flow reserve in humans," Circulation. 2003;107(14):1877-1883 Summary of the Invention [Means for solving the problem]

[0007] In embodiments herein, methods and systems are described for characterizing properties of microvascular tissue supplied with blood via a coronary artery under examination, the methods and systems comprising the following steps: i) acquiring a sequence of x-ray angiographic images of the coronary artery under examination, the images being acquired while a contrast agent is flowing into and passing through the coronary artery under examination; ii) determining a volumetric flow rate of flow through the coronary artery under examination using the angiographic image sequence of i); and iii) determining an index characteristic of the microvascular tissue supplied with blood via the coronary artery under test based on the volumetric flow rate of ii). Includes.

[0008] In an embodiment, operations i) through iii) may be performed automatically by a processor without human input.

[0009] In an embodiment, the volumetric flow rate can be based on the flow velocity of the contrast bolus tip within the angiographic image sequence of i) and the cross-sectional area of ​​the coronary artery under examination at multiple locations along the coronary artery under examination within the angiographic image sequence of i).

[0010] In an embodiment, the volumetric flow rate may be based on the transit time of the contrast bolus front within the angiographic image sequence of i) and the vascular volume of the coronary artery of interest.

[0011] In an embodiment, the vessel volume can be determined from a 3D reconstruction of the coronary artery of interest.

[0012] In an embodiment, the vascular volume may be based on determining one or more diameters of the coronary artery of interest along the coronary artery of interest.

[0013] In an embodiment, the flow velocity of the contrast bolus front can be determined from the distance traveled by the contrast bolus front as a function of time within the angiographic image sequence in i).

[0014] In an embodiment, the flow velocity of the contrast bolus tip can be determined from image analysis of the angiographic image sequence of i), which image analysis determines the proximal location of the coronary artery of interest, the distal location of the coronary artery of interest, the vascular path extending along the coronary vessel of interest from the proximal location to the distal location, and the propagation of the contrast bolus tip along the vascular path.

[0015] In embodiments, at least one of the proximal and distal locations may be determined using artificial intelligence and / or deep learning techniques.

[0016] In embodiments, artificial intelligence and / or deep learning techniques may use binary image segmentation.

[0017] In embodiments, the artificial intelligence and / or deep learning techniques may use additional information selected from the group consisting of the type of vessel, the rotation and angle used for image acquisition, ECG information, cardiac dominance information, and the time between image frames.

[0018] In an embodiment, the artificial intelligence and / or deep learning techniques may use a vascularity filter that is applied to multiple image frames of the angiographic image sequence of i).

[0019] In an embodiment, the proximal location can be determined from detecting the location of a guiding catheter used to inject contrast into the coronary vessel of interest.

[0020] In embodiments, the vascular path may be determined using a wave propagation algorithm between the proximal and distal locations.

[0021] In embodiments, the microvascular tissue may be part of the myocardium.

[0022] In an embodiment, the volumetric flow rate in ii) is characteristic of the volumetric flow rate during a portion of a cardiac cycle.

[0023] In an embodiment, ii) the volumetric flow rate is a property of the average flow velocity and the average volumetric flow rate over the cardiac cycle.

[0024] In an embodiment, the index may include quantitative data representative of the amount of dysfunction or resistance in the microvascular tissue supplied with blood via the coronary artery under test.

[0025] In an embodiment, the index may be determined from ii) the volumetric flow rate and a measurement of the pressure drop associated with the coronary artery under test.

[0026] In embodiments, the index may be normalized based on at least one parameter selected from the group consisting of heart weight, coronary volume, coronary artery cross-sectional area, patient weight, height, body surface area (BSA), or body mass index (BMI), cardiac dominance, or a combination thereof.

[0027] In an embodiment, the index may include quantitative data representing the ratio of flow through the coronary artery under test at rest to flow through the coronary artery under test in a hyperemic state.

[0028] In an embodiment, the method may include: i) determining a first volumetric flow rate of blood flow through the coronary artery under test while the patient is at rest using the at least one angiographic image; i) determining a second volumetric flow rate through the coronary artery under test while the patient is in an active / hyperemic state using the at least one angiographic image; and determining an index from the first and second volumetric flow rates.

[0029] In another aspect, a non-transitory computer-readable medium having stored thereon instructions that can be executed by causing a computing device to perform the methods described herein for characterizing properties of microvascular tissue supplied with blood via a coronary artery under test is provided.

[0030] In yet another aspect, an imaging system is provided that includes a data processor configured to perform the methods described herein to characterize properties of microvascular tissue supplied with blood via a coronary artery under examination.

[0031] Other aspects are described and claimed herein.

[0032] The features of the present invention and the advantages resulting therefrom will become more apparent from the following description of non-limiting embodiments thereof, as illustrated in the accompanying drawings. [Brief explanation of the drawings]

[0033] [Figure 1] 1 shows a flowchart of a method for determining microvascular dysfunction according to one embodiment of the present invention. [Figure 2] 1 illustrates a functional block diagram of an exemplary single-plane angiography system. [Figure 3A] 1 shows a functional block diagram of an exemplary method for calculating coronary volumetric flow rate based on contrast bolus velocity and vessel area. [Figure 3B] 3B is an example of an X-ray angiogram image that can be processed as part of the method of FIG. 3A to calculate coronary volumetric flow rate. [Figure 4A] 1 illustrates an exemplary method for calculating coronary flow velocity from an X-ray angiography image sequence. [Figure 4B] 4B is an example of an X-ray angiogram image that can be processed as part of the method of FIG. 4A to calculate coronary flow velocity. [Figure 5A] Tracking of the contrast bolus tip within an X-ray angiography image sequence (A1-A4) is shown. [Figure 5B1] 1 shows the coronary centerline of the vessel of interest after identifying its proximal and distal locations. [Figure 5B2] 1 shows an example of a tracked centerline of a vessel of interest. [Figure 6] 10 shows an example screenshot of a CAAS workstation bolus tracking in which the disclosed embodiments are implemented. [Figure 7] 1 shows a flowchart of a method for fully automatic initiation of contrast bolus tracking according to an embodiment of the present invention. [Figure 8] 1 shows an illustration of fully automatic centerline determination. [Figure 9] 10 shows an illustration of another method for determining coronary velocity. [Figure 10A] 1 shows a functional block diagram illustrating a method for calculating coronary volumetric flow rate based on contrast bolus propagation time and vessel volume. [Figure 10B]10B is an example of an X-ray angiogram image that can be processed as part of the method of FIG. 10A to calculate coronary volumetric flow rate. [Figure 11] An example of QCA (quantitative coronary analysis) 3D is shown. [Figure 12] A parabolic coronary velocity profile is shown. [Figure 13] 1 shows that coronary volumetric flow rate and flow velocity are not constant during the cardiac cycle. [Figure 14] 1 shows the pressure drop within the coronary artery system. [Figure 15] We demonstrate how to define myocardial mass or volume from an x-ray angiogram. [Figure 16] 1 shows another method for determining myocardial mass from an x-ray angiogram. [Figure 17] 1 shows an example of a high-level block diagram of an X-ray cine fluoroscopy system. [Figure 18] The American Heart Association defines the general model of the coronary tree. [Figure 19] 1 shows options for extracting different patient states and flows in different patient states. [Figure 20A] 10 illustrates an alternative approach for determining the propagation time of a contrast bolus. [Figure 20B] 20B is an example of an X-ray angiogram image that can be processed as part of the method of FIG. 20A to determine contrast bolus propagation time. [Figure 21A] 1 illustrates a different approach to assess myocardial status without determining contrast velocity or contrast bolus transit time. [Figure 21B] 21B is an example of an X-ray angiogram image that can be processed as part of the method of FIG. 21A to assess myocardial status. [Figure 22] The differences in microvascular cross-sections between normal microcirculation, structural microvascular dysfunction, and functional microvascular dysfunction are shown for resting and stressed microvessels. DETAILED DESCRIPTION OF THE INVENTION

[0034] The present disclosure describes methods and systems for assessing microvascular dysfunction using X-ray angiographic image data.

[0035] The present disclosure relates to methods and systems for quantifying microvascular dysfunction based on two-dimensional (2D) X-ray angiography image data, and will be disclosed primarily with reference to this field.

[0036] As used herein, the terms "image" or "image frame" refer to a single image, and the terms "image sequence" or "image data" can refer to multiple images acquired over time, and when used in connection with X-ray imaging, comprise multiple image frames covering one or more phases of the cardiac cycle.

[0037] 1 illustrates a flowchart illustrating operations according to one embodiment of the present application, which employ an imaging system capable of acquiring and processing one or more two-dimensional X-ray angiographic image sequences of a vascular organ (or portion thereof) or other object of interest. The one or more X-ray angiographic image sequences can be acquired, for example, using a single-plane or biplane angiography system. Examples of such systems include those manufactured by Siemens (Artis zee Biplane) or Philips (Allura Xper FD).

[0038] FIG. 2 is a functional block diagram of an exemplary single-plane angiography system. The system includes an angiography imager 212 that operates based on commands from a user interface module 216 and provides data to a data processing module 214. The single-plane angiography imager 212 captures two-dimensional X-ray image sequences of a vascular organ of interest, for example, in the anterior-posterior direction. The single-plane angiography imager 212 typically includes an X-ray source and detector pair mounted on the arms of a supporting gantry. The gantry positions the X-ray source and detector arms at various angles relative to a patient supported on a table between the X-ray source and detector. The data processing module 214 can be implemented by a personal computer, workstation, or other computer processing system. The data processing module 214 processes the two-dimensional image sequences captured by the single-plane angiography imager 212 to generate the data described herein. The user interface module 216 interacts with a user and communicates with the data processing module 214. The user interface module 216 can include various types of input / output devices, such as a display screen for visual output, a touch screen for touch input, a mouse pointer or other pointing device for input, a microphone for audio input, a speaker for audio output, a keyboard and / or keypad for input, etc. The data processing module 214 and the user interface module 216 cooperate to perform the operations of FIG. 1, as described below.

[0039] The operations of Figure 1 may also be performed by software code embodied in a computer product (e.g., an optical disk or other form of persistent memory such as a USB drive or a network server), or the software code may be loaded directly into the memory of a data processing system to perform the operations of Figure 1. Such a data processing system may also be physically separate from the angiography system used to acquire images as input using any type of data communication.

[0040] In this example, it is assumed that an X-ray imaging system acquires and stores at least one two-dimensional image sequence of the object of interest. Any imaging device capable of providing two-dimensional angiographic image sequences can be used for this purpose. For example, a biplane or single-plane angiographic system can be used. Examples of such systems are systems manufactured by Siemens (Artis zee Biplane) or Philips (Allura Xper FD).

[0041] An embodiment is disclosed with reference to Figure 1. The operations shown in Figure 1 may be performed in any logical order, and some may be omitted. Since the purpose of this application is to provide a selected (e.g., optimal) workflow that can be used during an intervention, reference will also be made to exemplary steps of the workflow.

[0042] As shown in Figure 1, the workflow comprises a number of steps. The first step (101) in Figure 1 is the acquisition of a patient-specific X-ray angiographic image sequence. Because the workflow of the present disclosure is based on filling the coronary vessels with contrast agent (or "contrast fluid"), the acquisition of angiographic image data can begin before the injection of contrast agent and continue during the filling of the coronary vessels of interest with contrast agent until the coronary vessels are completely filled with contrast agent and can also include flushing the contrast fluid out of the coronary vessels.

[0043] In step 102 of FIG. 1, coronary flow is derived from the X-ray image sequence acquired in step 101. As shown in the flowchart of FIG. 3A, the coronary volumetric flow rate (303) can be calculated by multiplying the flow velocity of the contrast bolus tip in the X-ray image sequence (301) by the cross-sectional area of ​​the vessel (302). The flow velocity of the contrast bolus in the X-ray image data will represent the coronary flow velocity. Coronary flow velocity can be derived from X-ray angiography as described in Non-Patent Document 6. For each frame during the contrast injection period, the distance traveled by the contrast agent (403) can be plotted against time (401), as shown in FIG. 4A. In FIG. 4A (401), the x-axis represents frames (time) within an X-ray angiographic image sequence, and the y-axis represents the distance from a proximal starting location (404) to the contrast bolus, e.g., shown at 405 in FIG. 4B, in a particular frame of the X-ray angiographic image sequence. Because the X-ray angiographic image sequence is acquired during contrast injection, later frames in the X-ray angiographic image sequence visualize more distal vessel locations (see also FIGS. 5A1-5A4), resulting in longer distances, as shown in graph 401. FIGS. 5A1-5A4 illustrate tracking the contrast bolus tip within an X-ray angiographic image sequence. Image 501 in FIG. 5A1 represents the starting frame. To begin bolus tracking, a centerline representing the coronary vessel segment of interest is required. This centerline can be derived, for example, from coronary artery segmentation as described in non-patent document 7, or can be manually identified. FIG. 5B1 shows coronary segmentation of the vessel of interest after identifying the proximal (507) and distal (508) locations at the starting frame 501 of FIG. 5A1, and the resulting centerline (509) after coronary segmentation. FIGS. 5A2-5A4 show tracking the contrast bolus tip backward in time, starting from the starting frame of FIG. 5A1. Image 504 of FIG. 5A2 is the previous frame in the image sequence relative to the starting frame (501). Image 505 of FIG. 5A3 is the previous frame in the image sequence relative to frame 504.Image 506 in Figure 5A4 is the previous frame in the image sequence relative to frame (505). Additionally, the centerline tracked from proximal position 507 in this sequence is shown as 510 in Figure 5A3 and 511 in Figure 5B2.

[0044] Referring to FIG. 4A, the ascending slope (402) of the distance traveled by the contrast bolus front plotted against time (401) represents the propagation of the contrast bolus. Coronary flow velocity can be determined by fitting a line (402), such as a straight line, to this ascending slope or by using the first derivative of the graph (401). FIG. 6 shows an example screenshot of bolus tracking in a Pie Medical Imaging CAAS Workstation. In this example, the above-described method is implemented within a QCA (Quantitative Coronary Artery Analysis) workflow (601); bolus tracking is performed after QCA segmentation, and a graph (602) is shown representing the distance traveled by the displaced contrast bolus front plotted against time. An initial centerline can be extracted from the QCA segmentation. If a 3D coronary artery reconstruction has already been performed, coronary velocities can be derived using both projections used to create the 3D coronary artery reconstruction. The 3D-based coronary velocities can be derived by calculating the average of the coronary velocities obtained by each projection, according to the method described in this patent application. A weighted average is also possible (the weights depend on the foreshortening of each projection used to create the 3D coronary artery reconstruction).

[0045] In a preferred embodiment, as described with reference to Figures 4A, 4B, 5A1-5A4, 5B1, 5B2, and 6, initiation of contrast bolus tracking is performed fully automatically without human intervention. This allows for the calculation of coronary blood flow and / or microvascular resistance immediately after acquiring X-ray angiographic image data according to the methods described in this patent application. This means that physicians have patient-specific coronary blood flow and / or microvascular resistance information immediately after image acquisition in the catheterization lab. Because multiple X-ray angiographic image acquisitions are performed during PCI and / or diagnostic coronary angiography, physicians can get a better impression of coronary blood flow and / or microvascular resistance during the procedure, thereby providing better treatment options. A method for fully automatic initiation of contrast bolus tracking is described below with reference to the flowchart of Figure 7.

[0046] In step 701, patient-specific X-ray image data is received. This is similar to step 101 in the flowchart of FIG. 1. Coronary angiography (also called cardiac catheterization or coronary arteriography) is a medical procedure used to diagnose and treat cardiovascular disease by imaging the coronary arteries that supply blood to the heart muscle. X-ray coronary angiography uses X-ray imaging technology to create detailed images of the coronary arteries. Contrast fluid is injected using a catheter placed into either the left or right coronary artery ostium. Injecting contrast fluid into the left coronary artery ostium images the LAD and LCX, while injecting contrast fluid into the right coronary artery ostium images the right coronary artery. To distinguish between the three major coronary vessels (RCA, LAD, LCX), in step 702 of the flowchart of FIG. 7, the type of coronary vessel is determined based on the image data (image sequence) of step 701. Step 702 is an optional step that can provide additional information to the physician (e.g., for reporting purposes) and additional input to the next workflow step in FIG. 7. The coronary vessel type can be determined by an artificial intelligence classification network trained to recognize the dominance of the RCA, LAD, and LCX coronary arteries in an x-ray image frame or sequence. Such a classification network can be developed using a deep learning network or a deep convolutional neural network. Deep learning techniques are described, for example, in Non-Patent Document 8. Optionally, the rotation and angle of the C-arm can also be used as additional information to improve the performance of such an artificial intelligence technique.

[0047] In step 703, the proximal start location is determined. The proximal start location represents the coronary artery ostium, which may be either the ostium of the right coronary artery or the ostium of the left coronary artery, depending on the contrast injection method, as described above. Figure 8 illustrates fully automated centerline determination. The proximal start location, designated "P" (801) in Figure 8, can be determined using, for example, artificial intelligence or deep learning techniques. Non-Patent Document 9 provides an example of a deep learning technique for determining the proximal start location (801). In the study, objects within an image are detected using bisection image segmentation. Binary image segmentation is a type of image segmentation technique that divides an image into two distinct regions or classes. The term "bisection" itself means dividing something into two parts. The goal of image segmentation is to divide an image into meaningful, homogeneous regions based on specific characteristics or criteria. The process of binary image segmentation typically involves distinguishing between two classes or regions in an image (often representing object and background, foreground and background, or different types of objects). Segmentation is performed based on specific features such as intensity, color, texture, and other visual characteristics. The proximal start location can be determined by training the IS-Net proposed by [9] to detect the entrances of the left and right coronary arteries. Additional inputs that can be used include information such as vessel type and / or C-arm rotation and angle (from step 702) to guide the neural network to landmarks that are optimal for a particular rotation and angle. Other information that can be provided includes information about the ECG, cardiac dominance, and the time between image frames. Another method for determining the proximal start location is to detect the guiding catheter in the image. When a guiding catheter is used to inject contrast into the left or right coronary artery, the tip of the guiding catheter indicates the proximal start location. An example of a method for detecting the tip of the guiding catheter is disclosed in Patent Document 1. This patent describes a method for detecting and tracking a catheter tip using a deep learning-based Bayesian filtering method.The method described in Patent Literature 1 models the likelihood term of Bayesian filtering with a convolutional neural network and comprehensively integrates it with particle filtering to achieve more robust catheter tip detection and tracking. In summary, for any position in an image, optical flow methods and noise addition can be used to predict the new position of the catheter tip (in a new image or a new frame in an image sequence) (to predict catheter movement). Furthermore, weights are updated by checking the likelihood of the position using a deep learning network. Next, all weights are normalized. The actual catheter tip position is equal to the weighted arithmetic mean of all positions and their weights. Finally, point resampling is performed around positions with high weight values. Alternatively, the proximal starting position can be determined by conventional image processing techniques, such as template matching using a catheter tip template or a coronary artery ostium template. Another method, proposed, for example, in Non-Patent Literature 10, involves running a vascularity filter on multiple frames in an image sequence. The output of the vascularity filter is then processed to determine the starting position, for example, using artificial intelligence techniques.

[0048] In step 704 of Figure 7, the distal location is determined. In Figure 8, an example of a distal location is shown at 802. The distal location can be detected automatically using artificial intelligence / deep learning. The artificial intelligence network can use an angiography image frame or an angiography image sequence as input. The network can detect landmarks distal to the coronary vessels. If multiple coronary vessels are visible in an image frame, the network can detect the distal location (802) for each. This would be the case in a left coronary angiogram, where the LAD and LCX are visible. An example of an artificial intelligence model that can detect these landmarks is IS-Net, proposed by [9]. Additional inputs that can be used are the type of vessel (from step 1402) and / or information such as the rotation or angle of the C-arm, which guides the neural network to the landmarks that are optimal for a particular rotation or angle. Other information that can be provided is information about the ECG, cardiac dominance, the time between image frames, etc. Another way to find the distal location (802) is to process the images directly. An example of finding the distal location is to run a vascular filter on multiple frames in an image sequence and track the change in the filter output. Such a filter has been proposed by

[10] .

[0049] In step 705 of FIG. 7 , the coronary vessel path (803) is determined using the proximal start location (result of step 703) and the distal location (result of step 704). The coronary vessel path can be determined, for example, using the determined proximal start location and distal location using CAAS Workstation 8.5, QCA workflow (Pie Medical Imaging, The Netherlands). Another method for determining the coronary vessel path is to use a wave propagation algorithm between the determined proximal start location and distal location. An example of a wave propagation algorithm is described in non-patent document 11. Another method for determining the coronary vessel path (803) is to first determine all vessels in the image and then select a segment between the proximal and distal markers. Such determination can be performed, for example, by a machine learning network proposed in non-patent document 12.

[0050] Returning to the description of 102 in FIG. 1, another method for determining coronary velocity, shown in FIG. 1, is described in U.S. Patent No. 5,623,995. In this method, the pixel intensity profile (FIG. 9, 903) along the centerline of a vessel (FIG. 9, 901) is analyzed in two temporally distinct image frames. The intensity changes at the contrast bolus front (FIG. 9, 902) and can be detected by the method disclosed in U.S. Patent No. 5,623,995. Coronary flow velocity can be determined using the length and time difference between the two frames.

[0051] To convert the coronary flow velocity (301) to the coronary volumetric flow rate (303), it must be multiplied by the cross-sectional area (302) shown in Figure 3A. This can be done by multiplying the cross-sectional area at each location along the vessel path (304 and 305) by the coronary flow velocity, as shown in Figure 3B, and integrating the result over many locations along the vessel length specified in equation (1). The advantage of integrating the cross-sectional area of ​​the vessel segment of interest is that, if there is a bifurcation, the decrease in coronary volumetric flow rate after bifurcation is taken into account by the decrease in cross-sectional area after bifurcation. The cross-sectional area can be determined, for example, by QCA3D, as described in detail in

[13] or

[14] , or by using densitometry or assuming circularity for the area calculation, as described in [7].

number

[0052] Coronary flow velocity can also be determined manually by dividing the length of the vessel segment by the time it takes for contrast to travel from the beginning (proximal; Figure 3, 304) to the end (distal; Figure 3, 305) of the vessel segment (contrast bolus propagation time). This length can be determined from an x-ray angiographic image, for example, using CAAS QCA3D or CAAS QCA. The time it takes for contrast to travel through the vessel segment can be determined by counting the number of frames it takes for the contrast to travel from the beginning (proximal) to the end (distal) position within the vessel segment of interest. This count can be converted to seconds by dividing the number of frames by the frame rate of the x-ray acquisition (e.g., 15 frames / second).

[0053] As shown in FIG. 10A, the volumetric flow rate of step 102 can also be derived by dividing the vessel volume (represented by block 1002) by the time it takes for a contrast bolus to travel from a proximal location (1004) to a distal location (1005) within the X-ray angiography image sequence shown in the exemplary image frame of FIG. 10B. This is referred to as the contrast bolus propagation time (in seconds), represented by block 1001 in FIG. 10A. Both the vessel volume (1002) and the contrast bolus propagation time (1001) must be determined for the same vessel segment, i.e., between the proximal location (1004) and the distal location (1005). The vessel volume (1002) can be determined, for example, using CAAS QCA3D, as shown in FIG. 11, based on a 3D reconstruction (1103) of the vessel using two or more X-ray angiography projection images with different viewing angles (1101) and (1102). Alternatively, vascular volume can be determined from a single two-dimensional X-ray angiographic image. For a single two-dimensional X-ray projection, the vessel area can be calculated using the diameter along the segmented artery (between 1004 and 1005) and converted to area using a densitometry method (e.g., using CAAS QCA (non-patent document 7)) in which the cross-sectional area is related to a gray value representing X-ray absorption. The vessel volume can be calculated by integrating the cross-sectional area over the path of the vessel (between 1004 and 1005). Alternatively, the contrast bolus propagation time can be determined by counting the number of frames it takes for the contrast agent to travel from a proximal to a distal location and dividing by the frame rate of the X-ray angiographic image sequence 101.

[0054] The contrast agent propagation time (1001 in FIG. 10A) can also be determined by converting the determined contrast agent bolus velocity (301 in FIG. 3) to time. This conversion is achieved by dividing the vessel length by the determined contrast agent bolus velocity (301). This results in the contrast agent propagation time. If the velocity is determined using two-dimensional images, the length can be derived from this two-dimensional image (e.g., performed using CAAS QCA) or obtained by QCA3D. If the velocity is determined using three-dimensional images (as described above), the length should also be determined in three-dimensional images, e.g., using CAAS QCA3D. The volumetric flow rate 102 or 1003 can be determined by dividing the previously determined vessel volume 1002 by this contrast agent propagation time 1001. This method is particularly important when the contrast agent bolus velocity (301) is determined using two-dimensional image information. This is because, in this situation, the foreshortening effect present in two-dimensional X-ray angiographic images can cause the determined velocity to be inaccurate. For example, this occurs when the contrast bolus velocity (301) is determined by an automated bolus tracking algorithm using two-dimensional image data, as described previously in this patent application. This foreshortening effect can be removed by converting the contrast bolus velocity (301) to a contrast bolus propagation time (1001) by dividing the vessel length by the determined contrast bolus velocity (301) based on the two-dimensional image data. Coronary blood flow can be calculated using this contrast bolus propagation time (1001) using the vessel volume (1002) based on three-dimensional information, for example, by using CAAS QCA3D (shown in FIG. 11 and designated 1002). Alternatively, vessel volume can be determined from a single two-dimensional X-ray angiographic image. For a single 2D x-ray projection, the vessel area can be calculated using the diameter along the segmented artery and converted to area by assuming the vessel is circular, or by using densitometry, where the cross-sectional area is related to a gray value representing x-ray absorption (e.g., by using CAAS QCA (non-patent document 7)).The vascular volume can be calculated by integrating the cross-sectional area over the length of the blood vessel.

[0055] The contrast agent solution has a higher viscosity (η) compared to blood and saline (saline is used, for example, during invasive IMR measurements). This viscosity difference can be corrected considering the flow of the contrast agent into the blood or saline. Assuming that the human body has the property of keeping the pressure difference constant, according to the Hagen-Poiseuille's law (Equation 2), when the length

Number

Number

Number

[0056] Assuming that the pressure drop is constant, the multiplication in Equation 2

Number

Number

Number

[0057] When determining contrast bolus velocity or contrast bolus transit time, it is important to consider whether the average velocity (or transit time) of all particles or the fastest particles are being measured. If the fastest particles are being measured, a correction may be required to convert the contrast bolus transit time or velocity of these fastest particles to the average transit time or average velocity of all particles. One example of this correction is to assume a Poiseuille profile, as shown in Figure 12, which has the property that the average velocity is half the maximum velocity within a parabolic velocity profile (1201).

[0058] During a cardiac cycle, the coronary volume flow rate and coronary flow velocity are not constant. As shown in FIG. 13 (diagram 1301), blood flow is low during myocardial contraction (systole) and high during myocardial relaxation (diastole). If the coronary volume flow rate and coronary flow velocity are measured over a portion of the cardiac cycle, the calculated coronary volume flow rate or coronary flow velocity can be corrected to the average coronary volume flow rate or average coronary flow velocity over the entire cardiac cycle. This can be done as described in U.S. Patent No. 6,233,999. In that patent application, the calculated velocity over a portion of the cardiac cycle is

number

number

number

number

number

number

number

number

[0059] Alternatively, the calculated coronary volumetric flow rate or coronary flow velocity

number

number

[0060] The above corrections are applied to the calculated coronary volumetric flow rate or coronary flow velocity

number

number

number

number

[0061] Flow through the myocardium can be adjusted based on the required blood supply by adjusting its resistance. Coronary blood flow should be reproducible and measured under the same patient conditions (stress levels). During the diastolic wave-free period (1302) of the cardiac cycle, microvascular resistance is naturally minimized without the need for hyperemia induced by the administration of vasodilators (15). This means that flow during the wave-free period (1908) and pharmacological vasodilation should be similar, and therefore the wave-free period can be used to determine flow using angiographic images acquired at rest with the same magnitude and variability. This wave-free period is within the diastolic phase of the cardiac cycle. Coronary flow in this patent application is determined by assessing the propagation of contrast fluid through the coronary arteries. To reliably measure blood flow during this period, contrast fluid injection must be initiated based on the electrocardiogram to ensure that contrast fluid passes through the coronary arteries of interest during this diastolic wave-free period. The potential delay between the start of contrast injection and the passage of contrast fluid into the vessel of interest can be taken into account. This can be determined for a patient population or estimated based on typical coronary blood flow, i.e., the velocity and distance traveled by the contrast fluid. Alternatively, blood flow velocity can be derived exclusively during the wave-free period, as described with reference to FIG. 3. This can be achieved by fitting a line (402) through only points within the wave-free period. Coronary blood flow (303) is then calculated according to Equation 1. The cross-sectional area used in Equation 1 is restricted to the vascular path traveled by the contrast bolus during the wave-free period. This is also true when deriving coronary flow according to the method described with reference to FIG. 10A, where the determination of contrast bolus propagation time (1001) must be restricted to the wave-free period, and vascular volume (1002) must be restricted to the vascular path corresponding to the path traveled by the bolus front during the wave-free period. The frames in the image sequence to which the waveless period applies can be determined by the ECG signal corresponding to the image data 101, for example as part of a DICOM file. If an ECG signal is not available, cardiac cycle information can be extracted as disclosed in US Pat. No. 6,239,999.

[0062] Contrast fluid induces hyperemia to some extent (Non-Patent Document 16). When an X-ray angiographic image of a patient is acquired while the patient is at rest (FIG. 19, 1901), there will be a hyperemic effect in the flow measurements caused by the contrast fluid used to image the coronary arteries using X-rays. This hyperemic effect reduces the variability of the flow measurements.

[0063] If flow is required for different patient states (e.g., resting and hyperemic) compared to the state at which the X-ray image data is acquired, the determined flow (1904) can optionally be corrected, e.g., from resting (1901) to hyperemic (1905), as shown in FIG. 19 . This can be done by modeling (e.g., polynomial approximation, machine learning, etc.) based on clinical data, such as invasive measurements of resting and hyperemic flows or X-ray image data, or by multiplying by a constant coefficient (e.g., 2.5) (Non-Patent Document 17) or by using a function such as that described in Non-Patent Document 18. Alternatively, an X-ray image of the patient during the hyperemic phase (1903) can be used to determine the hyperemic flow (1902). Calculation of microvascular resistance at various patient states (e.g., resting and hyperemic) is enabled by the modeling described in this section.

[0064] In step 103 of Figure 1, the pressure drop is determined.

number

number

number

[0065] Optionally, distal pressure can be measured invasively using pressure wires in the coronary arteries.

[0066] In step 104 of Figure 1, an index of microvascular dysfunction or resistance is calculated. In an embodiment, this index is quantitative data that represents the amount of dysfunction or resistance in the microvascular tissue of the myocardium supplied with blood through the coronary artery under test. One example of such an index is microvascular resistance, which is calculated by the ratio of the pressure drop in step 103 of Figure 1 to the volumetric flow rate determined in step 102 of Figure 1.

number

[0067] The index shown in Equation 5 is based on the existing index of microvascular resistance (IMR) determined from X-ray angiography (Non-Patent Documents 23, 24, and 25). Compared with these, the approach presented in this patent application is based on the contrast agent propagation time.

number

number

number

[0068] Optionally, in step 105, the index of microvascular dysfunction or microvascular resistance can be normalized, for example, to correct for inter-patient differences in myocardial blood supply and microvascular resistance. This normalization can be performed based on the following criteria: - Heart weight - Coronary volume - Coronary artery cross-sectional area - Patient's weight, height, body surface area (BSA), or body mass index (BMI) - Cardiac "dominance" - Combination: e.g., vascular-specific cardiac weight (combination of cardiac weight and cardiac dominance)

[0069] Normalization based on heart weight Myocardial weight or volume (heart weight, heart volume) affects the amount of blood supply required. The larger the myocardium, the greater the blood supply required. This blood supply is regulated by microvascular resistance, so there will be a correlation between heart weight or volume and microvascular resistance.

[0070] Myocardial mass or volume can be determined from an X-ray angiogram. One method is to segment the coronary arteries or coronary tree (1501) in one or more projections, as shown in Figure 15, providing information about the shape / geometry / size of the myocardium. The myocardial shape (1502) can be fitted to the shape of the coronary arteries (1503). When multiple projections are used, 3D reconstruction of the coronary arteries / tree can be used to more accurately determine the myocardial shape and, optionally, to model it in 3D.

[0071] Optionally, when using a single projection, the visible myocardial shape varies with projection angle, and the shape determined within a single projection can be corrected using prior knowledge of the myocardial shape based on the patient population and projection angle of the x-ray angiogram.

[0072] Optionally, if one coronary artery (e.g., the right coronary artery) is visible in the X-ray angiogram, the overall myocardial geometry can be estimated by combining prior knowledge of myocardial geometry based on the patient population and cardiac dominance (e.g., left dominance).

[0073] Another method for determining myocardial mass is to visualize the myocardial brush effect on the X-ray angiogram, as shown at 1601 in Figure 16. Alternatively, a predefined shape (e.g., an ellipse) can be fitted to the myocardium. When multiple projections are used, by using visualization or shape fitting from these multiple projection angles, the shape and mass of the myocardium can be determined more accurately, and optionally in 3D.

[0074] Optionally, this myocardium delineated within a single X-ray angiographic projection image can be refined based on prior knowledge of myocardial geometry based on the projection angle of the X-ray angiographic image and the patient population.

[0075] Another method for estimating myocardial mass is to use coronary computed tomography angiography (CCTA) image data. From CCTA image data, the myocardium can be visualized and segmented. Segmentation of the coronary arteries from CCTA image data can also be used to determine, guide, or estimate myocardial mass.

[0076] According to Non-Patent Document 27, the coronary volumetric flow rate given by Eq.

number

number

number

number

number

number

number

number

number

number

number

number

number

number

[0077] Alternatively, normalization can be performed by dividing the resistance by the myocardial weight or volume to obtain the resistance per unit of weight or volume (see Equation 10).

number

[0078] Instead, myocardial mass and microvascular resistance (myocardial resistance

number

number

number

number

number

number

number

number

number

[0079] Coronal volume-based normalization This normalization approach assumes that larger hearts (compared to the average) result in larger myocardial volumes and consequently a greater blood supply to the myocardium. Thus, coronary artery (tree) volume is proportional to the volumetric flow rate.

number

number

number

number

number

number

number

number

[0080] The scaling factors in Equations 14 and 15 can also be based on other relationships between the reference volume and the measured volume (eg, exponential, logarithmic, etc.).

[0081] Optionally, a relationship (formula / model) between coronary volume and flow can be determined based on a patient population. This relationship can be used to normalize the determined patient-specific flow to the flow belonging to a reference patient (e.g., a patient with a predefined coronary volume).

[0082] When using coronary volumes for normalization, the section of the coronary tree from which the volume should be calculated must be defined / specified to ensure that the same section is used for all patients. This section can be defined, for example, using a definition according to the American Heart Association's general model of the coronary tree, as shown in Figure 18.

[0083] Normalization based on coronal cross-sectional area Similar to coronary volume, the cross-sectional area of ​​the coronary arteries at specific locations (e.g., the left main, the ostium, or specific branches or bifurcations within the coronary tree) can also be used. Similar to the volumetric method, the scaling factor

number

number

number

number

number

number

number

[0084] The scaling factors in Equations 17 and 18 can also be other relationships between the reference and measured cross-sectional areas, such as exponential, logarithmic, etc.

[0085] Optionally, the cross-sectional area can also be used to determine a relationship (formula / model) between cross-sectional area and flow at a specified location based on population data, thereby normalizing the flow to that belonging to a reference patient (e.g., a patient with a given cross-sectional area at a specified location).

[0086] Normalization based on patient weight, height, BSA, or BMI A larger person will have a larger heart muscle that supplies blood to the entire body. Myocardial mass can be modeled using parameters such as the patient's weight, height, BSA, and BMI (e.g.,

number

number

number

number

[0087] Cardiac dominance-based normalization When blood flow is determined using any of the three major coronary arteries, i.e., the left anterior descending artery (LAD), the left circumflex artery (LCX), and the right coronary artery (RCA), the relative amount of total myocardial volume supplied by each coronary artery depends on cardiac dominance (i.e., left dominant, right dominant, or codominant). This must be taken into account to be able to compare measurements of vessels with different cardiac dominance. To achieve this, the flow determined in the x-ray angiography step 102 must be corrected. For example, assuming that the left dominant heart supplies a constant percentage of the myocardial mass (blood supply fraction (BSF)), the coronary volumetric flow rate should be corrected to 100% (see Equation 20). Using this corrected coronary volumetric flow rate, microvascular resistance can be calculated using Equation 21.

number

number

[0088] Combining normalization methods The above normalization techniques can be used in combination. For example, the weight of the myocardial portion supplied by a particular vessel (LAD, LCX, RCA, etc.) can be determined using cardiac dominance. This can be used to normalize resistance.

[0089] Reduced variability Because blood flow and pressure vary along the coronary arteries, variability can be reduced by determining blood flow and distal pressure at specific locations within the coronary tree (e.g., coronary artery pressure several centimeters after a particular branch).

[0090] (Alternative approach): Imaged blood flow as an index of microvascular dysfunction Microvascular resistance is derived from coronary blood flow and coronary blood pressure, as previously described. Critical to the normal functioning of the myocardium is an adequate oxygen supply by the coronary arteries and microvasculature. Coronary blood flow is a quantity directly related to the amount of oxygen delivered to the myocardium. Therefore, the calculated blood flow (step 102 in Figure 1) is: Microvascular dysfunctionIt can also be used as an index of risk or to predict future events such as cardiovascular death, myocardial infarction, hospitalization for heart failure, and ischemic revascularization. This method is particularly useful when the pressure drop in the coronary artery of interest is negligible or when patient-specific information required for accurate calculation of the pressure drop (e.g., resting aortic pressure when using vFFR) is missing and therefore incalculable. Coronary blood flow can optionally be normalized in a manner similar to that described in step 105 of Figure 1 to enable comparison of flow between patients. A coronary flow threshold can be used to distinguish between healthy coronary blood flow and reduced coronary blood flow due to, for example, microvascular dysfunction. There are several ways to define this coronary blood flow threshold. One example is based on the statistical difference in coronary blood flow between healthy and diseased populations. Diseased populations can be identified using invasive IMR or invasive bolus thermodilution. Another way to define diseased populations is to identify events after a predefined follow-up period, such as one year. This event can be defined as a major adverse cardiovascular event, which is a composite of cardiovascular death, myocardial infarction, hospitalization for heart failure, or ischemic revascularization. Based on coronary blood flow derived by the methods described in this patent application, the statistic used to define the threshold can be the lowest, middle, or highest tertile of coronary blood flow derived in the total population (healthy and diseased), or other statistical test to distinguish between the two groups. The above approach to distinguishing between healthy and reduced coronary blood flow can also be applied to the calculated microvascular resistance described in this patent application.

[0091] Improving currently established IMR measurements Quantitative X-ray image data analysis can improve the established IMR method for assessing microvascular disease (Non-Patent Document 28). The IMR calculated by Equation 22 is based on the assumption that vascular volume is constant. As mentioned above, vascular volume varies between patients. To overcome this assumption, the IMR index can correct for differences in vascular volume between patients by assessing vascular volume using, for example, CAAS QCA or CAAS QCA3D image analysis.

number

[0092] To account for vascular volume, Equation 22 can be substituted into Equation 23, where:

number

number

[0093] Alternatively, in the case of a stenosis, the coronary velocity increases due to the smaller cross-sectional area of ​​the stenosis, so the mean transit time

number

number

number

number

number

[0094] Alternatively, the distal pressure measurement in the established IMR measurement can be replaced by a non-invasive distal pressure calculation using geometric features of the vessel extracted from the X-ray image data, as performed, for example, within CAAS vFFR. These calculations can optionally be improved by incorporating the flow determined in step 102.

[0095] An alternative approach to contrast bolus propagation time As described in step 102, an alternative approach to determining the contrast bolus propagation time is by analyzing the density of contrast in both the proximal and distal coronary arteries, as illustrated in FIG. 20A. In the proximal coronary artery, a region of interest (ROI 2001) is shown, as illustrated in FIG. 20B. The mean, sum, median, or other index of pixel values ​​(called "density values") within this ROI can be determined for all frames of the x-ray angiography image sequence. These density values ​​can be plotted as a function of time (seconds) using the frame rate of image acquisition (2003, data points). The same plot can be created for the ROI labeled 2002 in the distal coronary artery, resulting in a graph represented by data points 2004. For both the proximal and distal ROIs (dotted lines 2003 and 2004), curves (e.g., polynomials) can be fit to the density values ​​over time. By determining the time difference of specific landmarks in the fitted curve, the propagation time of the contrast bolus can be obtained, for example, from the time difference between the peaks of the proximal and distal curves (as shown in 2005). Other landmarks, for example, the center of gravity of the area under the curve, the start of the ascending and / or descending slopes, can also be used. The "density value" mentioned above can also be the mean, sum, median, or other metric of the density measurement pixel values ​​within this ROI. Such density pixel values ​​can be obtained by background subtraction, for example, as described in non-patent document 7. Alternatively, the image sequence can be preprocessed to remove the background layer of the image sequence, for example, as described in non-patent document 29 or non-patent document 30. Optionally, cardiac and respiratory motion can be corrected, for example, using the method disclosed in patent document 6.

[0096] Alternative Approaches for Determining Indices of Microvascular Resistance or Dysfunction Another approach to assessing myocardial status without determining contrast velocity or contrast bolus transit time is to evaluate contrast density over time within a region of interest (ROI), as shown in FIG. 21A. If information about the overall myocardial status (whole myocardium) is desired, an ROI encompassing the entire myocardium (denoted 2101) can be shown, as shown in FIG. 21B. Evaluating the mean, sum, median, or other index of pixel values ​​within this ROI over time provides information about the rate of contrast outflow or inflow throughout this region. The time difference between the contrast outflow start point 2103 and the outflow end point 2104 provides insight into the rate at which contrast passes through the entire myocardium. This time is lower in healthy myocardium, where resistance is low. Additionally, the derivative of the downward slope 2105 provides information about the amount of contrast outflow and myocardial resistance. A similar approach can be applied to the upward slope, which represents inflow, or the time difference between the upward and downward slopes. Optionally, to make this time difference comparable between patients, the amount and injection rate of contrast agent can be standardized, or normalization can be performed based on, for example, the area under the curve using the plot of Figure 21 A. To determine microvascular resistance locally, the region of interest can be narrowed to the region of interest.

[0097] artificial intelligence Another approach to determining the index of microvascular status is to use artificial intelligence / deep learning. The artificial intelligence can use angiographic image data as input, but can also use additional information, such as the projection angle of the angiographic image, angiographic images from multiple projection angles, patient information such as age and weight, and even clinical information / data such as diabetes, hypertension, etc.

[0098] Artificial intelligence can be used to calculate an index indicative of microvascular status, or to calculate one or more of the aforementioned steps, for example, determining volumetric flow rate 102, determining pressure drop 103, or a combination thereof.

[0099] alternative index Another index providing information about the microvasculature is, for example, the coronary flow reserve (CFR) index, which is quantitative data representing the ratio of flow through the patient's coronary arteries when the patient is at rest to flow through the patient's coronary arteries when the patient is in an active / hyperemic state. In an embodiment, this parameter can be derived by determining the coronary volumetric flow rate based on images acquired while the patient is at rest, as described in step 102, or alternatively, based on images of the patient at rest and in a hyperemic state. Acquiring an X-ray angiographic image while the patient is in a hyperemic state is performed by, for example, administering adenosine or papaverine intracoronarily or intravenously to induce hyperemia, followed by X-ray acquisition of the coronary arteries, as described, for example, in non-patent document 31. The ratio of the coronary volumetric flow rate obtained from the X-ray angiographic image acquired at rest to the coronary volumetric flow rate obtained from the X-ray angiographic image acquired in a hyperemic state (or modeled from X-ray images acquired at rest) provides the CFR index, as shown in Equation 25:

number

[0100] Another index that provides information about microvascular dysfunction is the type of microvascular dysfunction. There are two types of microvascular dysfunction: structural and functional. Structural dysfunction involves physical changes or abnormalities in the microvasculature. This can include changes in vessel wall thickness, remodeling, or the presence of abnormalities such as fibrosis. Structural dysfunction can be the result of pathologies such as chronic inflammation, oxidative stress, and atherosclerosis. This results in reduced blood flow, increased resistance, and impaired nutrient exchange. Functional microvascular dysfunction refers to abnormalities in the dynamic regulation of blood flow and vascular reactivity, without necessarily involving physical changes in vascular structure, and can be due to impaired vasodilation (the inability of blood vessels to dilate appropriately) or impaired vasoconstriction (the inability of blood vessels to constrict appropriately). Figure 22 details the above-mentioned differences between normal microcirculation (2203), structural microvascular dysfunction (2204), and functional microvascular dysfunction (2205). Cross-sections of microvessels in the three aforementioned conditions are shown both at rest (2201) and under stress (2202). In normal microcirculation (2203), the resting vascular lumen (2207) is narrow, and its diameter is sufficient to supply sufficient oxygen to the myocardium. During stress, more oxygen is required, and the vascular tone of the blood vessels decreases, resulting in dilation of the vascular lumen. Vascular tone refers to the contractile activity of vascular smooth muscle cells in the walls of small arteries and arterioles (2206). In structural microvascular dysfunction (2204), the resting microvascular lumen functions similarly to the normal microcirculation, but vasodilation is limited during stress, resulting in a smaller lumen during stress compared to the normal microcirculation. Thus, in structural microvascular dysfunction, coronary blood flow and microvascular resistance are normal at rest, but coronary blood flow is reduced during stress, and microvascular resistance increases during stress. In functional microvascular dysfunction (2205), microvascular tone is already reduced at rest, and at times of stress, microvascular tone is generally similar or slightly reduced compared to normal microcirculation at times of stress.This means that in functional microvascular dysfunction, coronary blood flow at rest is normal (compensated for by vascular lumen expansion at rest) and microvascular resistance at rest is reduced, but during stress coronary blood flow also decreases and microvascular resistance during stress is similar to or slightly increased in normal microcirculation.

[0101] To distinguish between normal microcirculation (2203), structural microvascular dysfunction (2204), and functional microvascular dysfunction (2205), thresholds of coronary blood flow and / or microvascular resistance calculated according to the methods described herein based on X-ray angiographic image sequences at rest and / or during hyperemia are used. The latter can also be based on X-ray angiographic image sequences acquired at rest, in which hyperemia is simulated or modeled according to the methods described herein. There are several options for defining these thresholds. For example, they can be based on statistical differences in coronary blood flow and / or microvascular resistance between healthy and diseased populations (functional and / or structural microvascular dysfunction). Diseased populations and types of microvascular dysfunction can be identified by invasive IMR or invasive bolus thermodilution. Another way to define diseased populations is to identify events after a predefined follow-up period, such as one year. This event can be defined as a composite of major adverse cardiovascular events, cardiovascular death, myocardial infarction, hospitalization for heart failure, or revascularization due to ischemia. Using the coronary flow derived by the methods described in this patent application, the statistic used to define the threshold can be the lowest, middle, or highest tertile of derived coronary blood flow and / or microvascular resistance in the total population (healthy and diseased), or other statistical test to distinguish between the two groups.

[0102] The operations can be performed by a processor unit on a stand-alone system or a semi-stand-alone system that is connected to an X-ray cine-fluoroscopy system (FIG. 2) or other imaging system to acquire two-dimensional angiographic image sequences.

[0103] 17 shows an example of a high-level block diagram of an X-ray cine-fluoroscopy system, illustrating an example of how embodiments can be integrated into such a system.

[0104] Portions of the system (defined by the various functional blocks) may be implemented by dedicated hardware, analog and / or digital circuitry, and / or one or more processors operating program instructions stored in memory.

[0105] The x-ray system of Figure 17 includes an x-ray tube 1701 with a high voltage generator 1702 that generates an x-ray beam 1703. The high voltage generator 1702 controls and powers the x-ray tube 1701. The high voltage generator 1702 applies a high voltage to a vacuum gap between the cathode and rotating anode of the x-ray tube 1701. The voltage applied to the x-ray tube 1701 causes electrons to move from the cathode to the anode of the x-ray tube 1701, creating an x-ray photon generation effect, also known as bremsstrahlung. The generated photons form the x-ray beam 1703 that is directed toward an image detector 1706.

[0106] X-ray beam 1703 comprises photons with an energy spectrum ranging up to a maximum determined, among other things, by the voltage and current supplied to x-ray tube 1701. X-ray beam 1703 then passes through patient 1704, who is lying on adjustable table 1705. The x-ray photons of x-ray beam 1703 penetrate the patient's tissue to varying degrees. Different tissues of patient 1704 absorb different amounts of radiation, resulting in a modulation of the beam intensity. The modulated x-ray beam 1703 emerging from patient 1704 is detected by image detector 1706, located on the opposite side of the x-ray tube. This image detector 1706 can be either an indirect detection system or a direct detection system.

[0107] In an indirect detection system, the image detector 1706 comprises a vacuum tube (X-ray image intensifier) ​​that converts the X-ray emission beam 1703' into an amplified visible light image, which is transmitted to a visible light image receiver, such as a digital video camera, for image display and recording, resulting in a digital image signal.

[0108] In a direct detection system, image detector 1706 comprises a flat panel detector. The flat panel detector directly converts the x-ray exit beam 1703' into a digital image signal. The digital image signal output from image detector 1706 passes through digital image processing unit 1707. Digital image processing unit 1707 converts the digital image signal from 1706 into a corrected x-ray image (e.g., inverted and / or contrast-enhanced) in a standard image file format, such as DICOM. The corrected x-ray image can be stored on hard drive 1708.

[0109] 17 further comprises a C-arm 1709. The C-arm holds the X-ray tube 1701 and the image detector 1706 such that the patient 1704 and adjustable table 1705 are positioned between the X-ray tube 1701 and the image detector 1706. The C-arm can be moved (rotated and angulated) to a desired position to acquire a particular projection in a controlled manner using a C-arm control 1710. The C-arm control allows manual or automatic input to adjust the C-arm to a desired position for X-ray recording at a particular projection.

[0110] The X-ray system of Figure 17 can be either a single-plane imaging system or a bi-plane imaging system, in which case multiple C-arms 1709 are provided, each including an X-ray tube 1701, an image detector 1706, and a C-arm control 1710.

[0111] Additionally, the adjustment table 1705 can be moved using the table control unit 1711. The adjustment table 1705 can be moved along the x-axis, y-axis, and z-axis, and can also be tilted around a specific point.

[0112] Furthermore, the X-ray system is provided with a measurement unit 1713. This measurement unit contains information about the patient, such as ECG, aortic pressure, biomarkers, and / or height, length, etc.

[0113] The X-ray system is also provided with a general unit 1712 which can be used to interact with the C-arm control 1710, the table control 1711, the digital image processing unit 1707 and the measurement unit 1713.

[0114] In the x-ray system of Figure 17, an embodiment is implemented as follows: A clinician or other user acquires at least two x-ray angiographic image sequences of a patient 1704 by using a C-arm control 1710 to move a C-arm 1709 to a desired position relative to the patient 1704. The patient 1704 lies on an adjustable table 1705 that the user moves to a specific position using a table control 1711.

[0115] X-ray image sequences are then generated using high voltage generator 1702, X-ray tube 1701, image detector 1706, and digital image processing unit 1707, as described above. These images are then stored on hard drive 1708. Using these X-ray image sequences, general purpose processing unit 1712 performs the methods described in this application (e.g., as described in FIG. 1). As shown in FIG. 1, the workflow shown in FIG. 1 is performed using information from measurement unit 1713, digital image processing unit 1707, C-arm control unit 1710, and table control unit 1711.

[0116] Information obtained from the workflows described herein, including one or more indices characterizing the properties of the microvascular tissue (e.g., an index of microvascular tissue dysfunction or resistance and / or a coronary flow reserve (CFR) index), can be displayed on a display device, such as a display screen, operably coupled to the general processing unit 1712 of FIG. 17.

[0117] Several embodiments of methods and apparatus for restoring missing information regarding the order of velocity components and the direction of flow have been described and illustrated herein. While specific embodiments of the invention have been described, the invention is not limited thereto, and it is intended that the scope of the invention be as broad as the art will permit, and that the specification be so interpreted. For example, data processing operations can be performed offline on images stored in digital storage, such as PACS commonly used in the medical imaging field. Accordingly, those skilled in the art will recognize that still other modifications can be made to the invention without departing from the spirit and scope of the invention as defined by the appended claims.

[0118] The embodiments described herein may include various data stores and other memory and storage media, as discussed above. These may reside in a variety of locations, such as on storage media local to (and / or resident on) one or more computers, or on storage media remote from any or all computers across a network. In certain embodiments, information may reside on a storage area network (SAN), as known to those skilled in the art. Similarly, files necessary to perform functions attributed to a computer, server, or other network device may be stored locally and / or remotely as needed. When a system includes computerized devices, each device may include hardware elements that may be electrically coupled via a bus. These elements may include, for example, at least one central processing unit ("CPU" or "processor"), at least one input device (e.g., a mouse, keyboard, controller, touchscreen, or keypad), and at least one output device (e.g., a display device, printer, or speaker). Such a system may also include one or more storage devices, such as disk drives, optical storage devices, and solid-state storage devices such as random access memory (“RAM”) or read-only memory (“ROM”), as well as removable media devices, memory cards, flash cards, and the like.

[0119] Such devices may also include computer-readable storage media readers, communication devices (e.g., modems, network cards (wireless or wired), infrared communication devices, etc.), and the working memory described above. The computer-readable storage media reader is configured to connect to or receive computer-readable storage media representing remote, local, fixed, and / or removable storage devices, as well as storage media for temporarily and / or more permanently containing, storing, transmitting, and reading computer-readable information. The system and various devices will typically also include numerous software applications, modules, services, or other elements located within at least one working memory device, including an operating system and application programs such as client applications or web browsers. It should be understood that alternative embodiments may have various variations of the above. For example, customized hardware may be used, or particular elements may be implemented in hardware, software (including portable software such as applets), or both. Additionally, connections to other computing devices, such as network input / output devices, may be used.

[0120] Various embodiments may also receive, transmit, or store instructions and / or data implemented in accordance with the foregoing description on computer-readable media. Storage media and computer-readable media for storing code or portions of code may include any suitable media known or used in the art. This includes, but is not limited to, storage and communication media implemented in any method or technology for storage and / or transmission of information such as computer-readable instructions, data structures, program modules, or other data, including volatile and nonvolatile, removable and non-removable media. This includes RAM, ROM, electrically erasable programmable read-only memory ("EEPROM"), flash memory or other memory technology, compact disc read-only memory ("CD-ROM"), digital versatile disc (DVD) or other optical storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage, or any other medium that can be used to store the desired information and that can be accessed by a system device. Those skilled in the art will recognize other ways and / or methods for implementing the various embodiments based on the disclosure and teachings herein.

[0121] Accordingly, the specification and drawings are to be regarded in an illustrative rather than a restrictive sense, although it will be evident that various variations and modifications may be made thereto without departing from the broader spirit and scope of the invention as set forth in the appended claims.

[0122] Other variations are within the spirit of the present disclosure. Thus, while the disclosed technology is susceptible to various modifications and alternative constructions, specific embodiments thereof are shown in the drawings and have been described above in detail. It should be understood, however, that the application is not intended to limit the invention to the particular forms disclosed, but rather to cover all modifications, alternative constructions, and equivalents included within the spirit and scope of the invention as defined by the appended claims.

[0123] Use of the terms "a," "an," and "the," and similar referents in the context of describing the disclosed embodiments (particularly in the context of the claims below) should be construed to cover both the singular and the plural, unless otherwise indicated herein or clearly contradicted by context. The terms "comprising," "having," "including," and "containing" should be construed as open-ended (i.e., meaning "including, but not limited to"), unless otherwise noted. The term "connected," when unmodified and referring to a physical connection, should be construed as partially or wholly encompassed, connected, or coupled, even if there is something intervening. Recitation of ranges of values ​​herein is intended merely as a shorthand method of individually referring to each individual value falling within the range, unless otherwise stated herein, and each individual value is incorporated herein as if it were individually set forth herein. Use of the term "set" (e.g., "set of items") or "subset" should be construed as a non-empty set containing one or more members, unless otherwise stated or contradicted by context. Furthermore, unless otherwise stated or contradicted by context, the term "subset" of a corresponding set does not necessarily mean a proper subset of the corresponding set, and a subset and a corresponding set may be equivalent.

[0124] The operations of the processes described herein can be performed in any suitable order unless otherwise indicated herein or clearly contradicted by context. The processes described herein (or variations and / or combinations thereof) can be executed under the control of one or more computer systems configured with executable instructions and implemented as code (e.g., executable instructions, one or more computer programs, or one or more applications) collectively executed on one or more processors, by hardware, or by a combination thereof. The code can be stored on a computer-readable storage medium, for example, in the form of a computer program including a plurality of instructions executable by one or more processors. The computer-readable storage medium can be non-transitory.

[0125] Preferred embodiments of the present disclosure are described herein, including the best mode known to the inventors for carrying out the invention. Variations of these preferred embodiments will become apparent to those skilled in the art upon reading the foregoing description. The inventors expect skilled artisans to employ such variations as appropriate, and the inventors intend for the embodiments of the present disclosure to be practiced differently from that specifically described herein. Accordingly, the scope of the present disclosure includes all variations and equivalents of the subject matter recited in the claims appended hereto as permitted by applicable law. Furthermore, any combination of all possible variations of the above-described elements is encompassed within the scope of the present disclosure unless otherwise indicated herein or clearly contradicted by context.

[0126] All references cited in this specification, including publications, patent applications, and patents, are hereby incorporated by reference to the same extent as if each reference was individually and specifically indicated to be incorporated by reference and was set forth in its entirety herein.

Claims

1. 1. A computer-implemented method for characterizing properties of microvascular tissue supplied with blood via a coronary artery under examination from a sequence of x-ray angiographic images of the coronary artery under examination acquired while a contrast agent flows into and through the coronary artery, the method comprising: i) determining a volumetric flow rate of flow through the coronary artery under test using the angiographic image sequence; and ii) determining an index characteristic of the microvascular tissue supplied with blood via the coronary artery under examination based on the volumetric flow rate; A method for providing the above.

2. the volumetric flow rate is based on a flow velocity of a contrast bolus front within the angiographic image sequence and a cross-sectional area of ​​the coronary artery under examination at a plurality of locations along the coronary artery under examination within the angiographic image sequence. The method of claim 1.

3. the volumetric flow rate is based on a transit time of a contrast agent bolus front within the angiographic image sequence and a vascular volume of the coronary artery of interest. The method according to claim 1 or 2.

4. the vascular volume is determined from a 3D reconstruction of the coronary artery of interest; The method of claim 3.

5. the vascular volume is based on determining one or more diameters of the coronary artery of interest along the coronary artery of interest; 5. The method according to claim 3 or 4.

6. the flow velocity of the contrast bolus front is determined from the distance traveled by the contrast bolus front as a function of time within the angiographic image sequence; The method according to any one of claims 1 to 5.

7. the flow velocity of the contrast bolus front is determined from image analysis of the angiographic image sequence, the image analysis determining a proximal location of the coronary artery of interest, a vascular path extending along the coronary vessel of interest from the proximal location to the distal location, and a propagation of the contrast bolus front along the vascular path; The method according to any one of claims 1 to 6.

8. At least one of the proximal location and the distal location is determined using artificial intelligence and / or deep learning techniques. The method of claim 7.

9. wherein the artificial intelligence and / or deep learning techniques use binary image segmentation; The method of claim 8.

10. wherein the artificial intelligence and / or deep learning techniques use additional information selected from the group consisting of: vessel type, rotation and angle used for image acquisition, ECG information, cardiac dominance information, and time between image frames; 10. The method according to claim 8 or 9.

11. the artificial intelligence and / or deep learning techniques use a vascularity filter applied to a plurality of image frames of the angiographic image sequence; The method according to any one of claims 8 to 10.

12. the proximal location is determined from detecting the position of a guiding catheter used to inject the contrast agent into the coronary vessel of interest. The method according to any one of claims 7 to 11.

13. the vascular path is determined using a wave propagation algorithm between the proximal and distal locations; The method according to any one of claims 7 to 12.

14. the microvascular tissue is part of the myocardium; The method according to any one of claims 1 to 13.

15. ii) the volumetric flow rate is characteristic of a volumetric flow rate for a portion of a cardiac cycle; The method according to any one of claims 1 to 14.

16. the volumetric flow rate being a characteristic of average flow velocity and average volumetric flow rate over the cardiac cycle; The method according to any one of claims 1 to 15.

17. the index comprises quantitative data representative of the amount of dysfunction or resistance in the microvascular tissue supplied with blood via the coronary artery under test; The method according to any one of claims 1 to 16.

18. determining a pressure drop associated with the coronary artery under test; ii) the index is determined from the volumetric flow rate and the pressure drop of i); 18. The method of claim 17.

19. ii) the index is normalized based on at least one parameter selected from the group consisting of heart weight, coronary volume, coronary artery cross-sectional area, patient weight, height, body surface area (BSA), or body mass index (BMI), cardiac dominance, or a combination thereof; 19. The method of claim 17 or 18.

20. the index comprises quantitative data representing a ratio of flow through the coronary artery under test at rest to flow through the coronary artery under test in the hyperemic state; 20. The method according to any one of claims 1 to 19.

21. determining a first volumetric flow rate of blood flow through the coronary artery under test while the patient is at rest using the at least one angiographic image; determining a second volumetric flow rate of blood flow through the coronary artery under test in the patient's active / hyperemic state using the at least one angiographic image; and determining the index from the first and second volumetric flow rates; 21. The method of claim 20, comprising:

22. A computer readable non-transitory medium storing instructions that, when executed by a computing device, cause the computing device to perform the method of any one of claims 1 to 21.

23. 22. An apparatus for acquiring an image dataset of a patient, the apparatus comprising a data processing module configured to carry out the method according to any one of claims 1 to 21 to characterize properties of the microvascular tissue supplied with blood via a coronary artery under examination.

Images