Systems, methods, and devices for medical image analysis, diagnosis, severity classification, decision making, and / or disease tracking - Patents.com

Abstract

The disclosure herein relates to systems, methods, and devices for medical image analysis, diagnosis, severity classification, decision making, and / or disease tracking. In some embodiments, the systems, devices, and methods described herein are configured to analyze non-invasive medical images of a subject to automatically and / or dynamically identify one or more features, such as plaque and vessels, and / or derive one or more quantified plaque parameters, such as radiodensity, radiodensity composition, volume, radiodensity heterogeneity, geometry, location, and the like, perform computational fluid dynamics analysis, facilitate assessment of heart disease and coronary artery disease risk, improve drug development, determine CAD risk factor targets, provide characterization of atherosclerosis and vessel morphology, determine an indication of myocardial risk, and / or the like. In some embodiments, the systems, devices, and methods described herein are further configured to generate one or more assessments of plaque-based disease from the raw medical images using one or more of the identified features and / or quantified parameters.

Description

[Technical field]

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS This application is a continuation-in-part of U.S. Provisional Patent Application No. 17 / 662,734, filed May 10, 2022. This application further claims the benefit of U.S. Provisional Patent Application No. 63 / 235,010, filed August 19, 2021, U.S. Provisional Patent Application No. 63 / 241,427, filed September 7, 2021, U.S. Provisional Patent Application No. 63 / 276,268, filed November 5, 2021, U.S. Provisional Patent Application No. 63 / 264,805, filed December 2, 2021, U.S. Provisional Patent Application No. 63 / 264,913, filed December 3, 2021, and U.S. Provisional Patent Application No. 63 / 296,116, filed January 3, 2022. U.S. Patent Application No. 17 / 662,734 is a continuation of U.S. Patent Application No. 17 / 367,549, filed July 5, 2021, a continuation of U.S. Patent Application No. 17 / 350,836, filed June 17, 2021, a continuation-in-part of U.S. Patent Application No. 17 / 213,966, filed March 26, 2021, a continuation of U.S. Patent Application No. 17 / 142,120, filed January 5, 2021, and claims the benefit of U.S. Provisional Patent Application No. 62 / 958,032, filed January 7, 2020. U.S. Patent Application No. 17 / 350,836 claims the benefit of U.S. Provisional Patent Application No. 63 / 201,142, filed April 14, 2021, U.S. Provisional Patent Application No. 63 / 041,252, filed June 19, 2020, U.S. Provisional Patent Application No. 63 / 077,044, filed September 11, 2020, U.S. Provisional Patent Application No. 63 / 077,058, filed September 11, 2020, U.S. Provisional Patent Application No. 63 / 089,790, filed October 9, 2020, and U.S. Provisional Patent Application No. 63 / 142,873, filed January 28, 2021. Each and every one of the disclosures cited above is incorporated herein by reference in its entirety. Also, U.S. Patent No. 10,813,612 is incorporated herein by reference in its entirety. Any and all applications for which foreign or domestic priority is identified in an Application Data Sheet filed with this application are hereby incorporated by reference pursuant to 37 CFR §1.57.

[0002] The present application relates to systems, methods, and devices for medical image analysis, diagnosis, severity classification, decision making, and / or disease tracking. [Background technology]

[0003] Coronary heart disease affects over 17.6 million Americans. The current trend in treating cardiovascular health issues is nearly two-fold. First, physicians typically examine the patient's cardiovascular health at a macro level, for example, by analyzing biochemistry or blood content or biomarkers to determine whether there are high levels of cholesterol elements in the patient's bloodstream. In response to high levels of cholesterol, some physicians prescribe one or more drugs, such as statins, as part of a treatment plan to reduce the perceived high levels of cholesterol elements in the patient's bloodstream.

[0004] A second general trend for currently treating cardiovascular health problems involves physicians assessing a patient's cardiovascular health by using angiograms to identify large blockages in the patient's various arteries. In response to finding large blockages in various arteries, the physician sometimes performs an angioplasty procedure in which a balloon catheter is guided to the point where the blood vessel is narrowed. Once properly positioned, the balloon is inflated to compress or flatten plaque or fatty material into the artery wall and / or stretch the artery open to increase blood flow through the vessel and / or to the heart. Sometimes, the balloon is used to position and expand a stent within the vessel to compress plaque and / or keep the vessel open to allow more blood to flow. Approximately 500,000 cardiac stent procedures are performed in the United States each year.

[0005] However, a recent $100 million federally funded study questions whether current trends in cardiovascular disease treatment are the most effective treatment for all types of patients. The recent study included over 5,000 patients with moderate to severe stable heart disease from 320 locations in 37 countries, and provided new evidence that stent and bypass surgery procedures are unlikely to be more effective than drugs combined with lifestyle changes in patients with stable heart disease. Therefore, it may be advantageous for patients with stable heart disease to forgo invasive surgical procedures such as angioplasty and / or cardiac bypass and instead receive prescriptions for heart medications such as statins and make certain lifestyle changes, such as regular exercise. This new treatment plan could impact thousands of patients worldwide. It is estimated that one-fifth of the estimated 500,000 cardiac stent procedures performed annually in the United States are for patients with stable heart disease. Furthermore, it is estimated that 25% of the estimated 100,000 patients with stable heart disease, approximately 23,000 people, do not experience chest pain. Thus, over 20,000 patients annually may forego invasive surgical procedures or may avoid complications as a result of such procedures.

[0006] It may be important to better understand a patient's cardiovascular disease in order to determine whether the patient should forgo invasive surgical procedures and instead opt for a medical treatment plan. In particular, it may be advantageous to have a better understanding of the health of the patient's arterial blood vessels. Summary of the Invention

[0007] Various embodiments described herein relate to systems, methods, and devices for medical image analysis, diagnosis, severity classification, decision making, and / or disease tracking.

[0008] In particular, in some embodiments, the systems, devices, and methods described herein are configured to identify one or more coronary arteries and / or plaque therein utilizing non-invasive medical imaging techniques, such as, for example, CT images, which may be input into a computer system configured to automatically and / or dynamically analyze medical images. For example, in some embodiments, the system may be configured to automatically and / or dynamically analyze medical images utilizing one or more machine learning and / or artificial intelligence algorithms to identify, quantify, and / or classify one or more coronary arteries and / or plaque. In some embodiments, the system may be further configured to utilize the identified, quantified, and / or classified one or more coronary arteries and / or plaque, for example, using one or more artificial intelligence and / or machine learning algorithms, to generate treatment plans, track disease progression, and / or provide medical reports regarding patient characteristics. In some embodiments, the system may be further configured to dynamically and / or automatically generate visualizations of the identified, quantified, and / or classified one or more coronary arteries and / or plaque, for example, in the form of a graphical user interface. Further, in some embodiments, to calibrate medical images obtained from different medical imaging scanners and / or different scanning parameters or environments, the system can be configured to utilize a normalization device comprising one or more sections of one or more substances.

[0009] In some embodiments, a normalization device configured to normalize medical images of a subject's coronary artery region for algorithm-based medical imaging analysis comprises at least a substrate configured with a size and shape to be placed on a medical imaging device along with a patient such that the normalization device and the patient can be imaged together such that a region of interest of the patient and the normalization device appear in a medical image taken by the medical imaging device; a plurality of compartments positioned on or within the substrate, the arrangement of the plurality of compartments being fixed on or within the substrate; and a plurality of samples positioned within one of the plurality of compartments, each having a known volume, absolute density, and relative density, the plurality of samples including a set of contrast samples, each of the contrast samples having an absolute density different from the other absolute densities of the contrast samples, a set of calcium samples, each of the calcium samples having an absolute density different from the other absolute densities of the calcium samples, and a set of fat samples, each of the fat samples having an absolute density different from the other absolute densities of the fat samples, the contrast samples of the set being positioned within the plurality of compartments such that the set of calcium samples and the set of fat samples surround the set of contrast samples.

[0010] In some embodiments, the normalization device further comprises an attachment mechanism disposed on the substrate, the attachment mechanism configured to attach the normalization device to the patient such that the normalization device and the patient can be imaged together such that regions of interest of the patient and the normalization device appear in a medical image taken by the medical imaging device. In some embodiments of the normalization device, the set of contrast samples includes four contrast samples, the set of calcium samples includes four calcium samples, and the set of fat samples includes four fat samples. In some embodiments of the normalization device, the plurality of samples further includes at least one of an air sample and a water sample. In some embodiments of the normalization device, a volume of the first contrast sample is different from a volume of the second contrast sample, a volume of the first calcium sample is different from a volume of the second calcium sample, and a volume of the first fat sample is different from a volume of the second fat sample. In some embodiments of the normalization device, the first contrast sample is disposed within the plurality of compartments such that it is adjacent to the second contrast sample, the first calcium sample, and the first fat sample. In some embodiments of the normalization device, the first calcium sample is disposed within the plurality of compartments adjacent to the second calcium sample, the first contrast sample, and the first fat sample. In some embodiments of the normalization device, the first fat sample is disposed within the plurality of compartments adjacent to the second fat sample, the first contrast sample, and the first calcium sample. In some embodiments of the normalization device, the set of contrast samples, the set of calcium samples, and the set of fat samples are disposed in a manner that mimics a blood vessel.

[0011] In some embodiments, a computer-implemented method for generating an atherosclerotic cardiovascular disease (ASCVD) risk assessment using a normalization device, where normalization of medical imaging improves accuracy of algorithm-based imaging analysis, the method includes receiving a first set of images of a first arterial bed and a first set of images of a second arterial bed, where the second arterial bed is non-contiguous with the first arterial bed, and where at least one of the first set of images of the first arterial bed and the first set of images of the second arterial bed are normalized using the normalization device; quantifying ASCVD in the first arterial bed using the first set of images of the first arterial bed; quantifying ASCVD in the second arterial bed using the first set of images of the second arterial bed; and determining a first ASCVD risk score based on the quantified ASCVD in the first arterial bed and the quantified ASCVD in the second arterial bed.

[0012] In some embodiments, the method for generating an atherosclerosis cardiovascular disease (ASCVD) risk assessment further includes determining a first weighted assessment for a first arterial bed based on the quantified ASCVD for the first arterial bed and the weighted adverse events for the first arterial bed, and determining a second weighted assessment for a second arterial bed based on the quantified ASCVD for the second arterial bed and the weighted adverse events for the second arterial bed, and determining the first ASCVD risk score further includes determining an ASCVD risk score based on the first weighted assessment and the second weighted assessment. Further, in some embodiments, the method for generating an atherosclerosis (ASCVD) risk assessment further includes receiving a second set of images of a first arterial bed and a second set of images of a second arterial bed, where the second set of images of the first arterial bed is generated after generating the first set of images of the first arterial bed and the second set of images of the second arterial bed is generated after generating the first set of images of the second arterial bed, quantifying ASCVD in the first arterial bed using the second set of images of the first arterial bed, quantifying ASCVD in the second arterial bed using the second set of images of the second arterial bed, and determining a second ASCVD risk score based on the quantified ASCVD in the first arterial bed using the second set of images and the quantified ASCVD in the second arterial bed using the second set of images. In some embodiments of the method for generating an atherosclerosis (ASCVD) risk assessment, determining the second ASCVD risk score is further based on the first ASCVD risk score. In some embodiments of the method for generating an atherosclerosis (ASCVD) risk assessment, the first arterial bed includes one of the aorta, the carotid artery, the arteries of the lower extremities, the renal artery, or the cerebral artery, and the second arterial bed includes one of the aorta, the carotid artery, the arteries of the lower extremities, the renal artery, or the cerebral artery that is different from the arteries of the first arterial bed.

[0013] In some embodiments, a computer-implemented method for generating a multimedia medical report for a patient based on images generated using a normalization device, where the normalization device improves accuracy of non-invasive medical image analysis, the medical report being associated with one or more tests for the patient. The method includes receiving an input of a request to generate a medical report for the patient, the request indicating a format of the medical report; receiving patient information for the patient, where the patient information is associated with the report generation request; determining one or more patient characteristics associated with the patient using the patient information; and accessing an association between a type of medical report and the patient medical information, where the patient medical information indicates medical images for the patient and test results of one or more tests performed on the patient. and a medical image is generated using the normalization device; accessing report content associated with the patient's medical information and the requested medical report, the report content including multimedia content not related to a particular patient, the multimedia content including a greeting segment in the patient's language, an explanation segment describing the type of test performed, a result segment to communicate the test results, and an explanation segment describing the results of the test, and a conclusion segment, at least a portion of the multimedia content including test results and one or more medical images associated with the tests performed on the patient; and generating the requested medical report using the patient information and the report content based at least in part on a format of the medical report.

[0014] In some embodiments, the normalization device is used to assess coronary artery disease (CAD) in a subject by generating one or more CAD risk scores for the subject based on multi-dimensional information derived from non-invasive medical image analysis.1. A computer-implemented method for assessing risk of coronary artery disease, wherein a normalization device improves accuracy of non-invasive medical image analysis, the method comprising: accessing, by a computer system, medical images of a coronary artery region of a subject, the medical images of the coronary artery region of the subject being non-invasively acquired; identifying, by the computer system, one or more segments of a coronary artery in the medical images of the coronary artery region of the subject; and determining, by the computer system, one or more plaque parameters, vascular parameters, and clinical parameters for each of the identified one or more segments of the coronary artery, the one or more plaque parameters comprising one or more of plaque volume, plaque composition, plaque attenuation, or plaque location; the one or more vascular parameters comprising one or more of stenosis severity, lumen volume, percentage of coronary blood volume, or percentage of fractional myocardial mass; and the one or more clinical parameters comprising one or more of age percentile health status or gender percentile health status. generating, by the computer system, a weighted measure of the determined one or more plaque parameters, vascular parameters, and clinical parameters for each of the identified one or more segments of the coronary artery, wherein the weighted measure is generated by applying a correction factor; combining, by the computer system, the generated weighted measures of the determined one or more plaque parameters, vascular parameters, and clinical parameters for each of the identified one or more segments of the coronary artery to generate a CAD risk score for each of the one or more vessels, vascular regions, or subject; and generating, by the computer system, a graphical plot of the generated one or more per-vessel, per-vascular region, or per-subject CAD risk scores for visualizing and quantifying the subject's risk of CAD based on one or more of the per-vessel, per-vascular region, or per-subject, wherein the computer system includes a computer processor and an electronic storage medium.

[0015] In some embodiments, a computer-implemented method for tracking efficacy of a medical treatment for a plaque-based disease based on non-invasive medical image analysis using a normalization device, where the normalization device improves accuracy of the non-invasive medical image analysis, the method includes accessing, by a computer system, a first set of plaque parameters and a first set of vascular parameters associated with a subject, where the first set of plaque parameters and the first set of vascular parameters are derived from a first medical image of the subject including one or more regions of plaque, where the first medical image of the subject is non-invasively acquired at a first time point, where the first set of plaque parameters include one or more of a density, a location, or a volume of the one or more regions of plaque from the medical image of the subject at the first time point, where the first set of vascular parameters includes vascular remodeling of the vascular structure at the first time point; and accessing, by the computer system, a second medical image of the subject, where the second medical image of the subject is non-invasively acquired at a second time point after the subject has been treated with the medical procedure. accessing, by the computer system, a second medical image of the subject, the second time point being later than the first time point, the second medical image including one or more regions of plaque; identifying, by the computer system, the one or more regions of plaque from the second medical image; and determining, by the computer system, a second set of plaque parameters and a second of vascular parameters associated with the subject by analyzing the one or more regions of plaque from the second medical image, the second set of plaque parameters including one or more of a density, a location, or a volume of the one or more regions of plaque from the medical image of the subject at the second time point, the second set of vascular parameters including vascular remodeling of the vascular structure at the second time point; and analyzing, by the computer system, one or more changes between the first set of plaque parameters and the second set of vascular parameters;tracking a progression of the plaque-based disease based on one or more of the analyzed changes between the first set of plaque parameters and the second set of plaque parameters, or the analyzed changes between the first set of vascular parameters and the second set of vascular parameters, and determining, by the computer system, efficacy of a medical treatment based on the tracked progression of the plaque-based disease, the computer system including a computer processor and an electronic storage medium.

[0016] In some embodiments, a computer-implemented method for determining a continuing personalized treatment for a subject at risk for atherosclerosis (ASCVD) based on coronary CT angiography (CCTA) analysis using one or more quantitative imaging algorithms with a normalization device, where the normalization device improves accuracy of the one or more quantitative imaging algorithms, includes: assessing, by the computer system, a baseline ASCVD risk of the subject by analyzing a baseline CCTA analysis result using the one or more quantitative imaging algorithms, where the baseline CCTA analysis result is based at least in part on one or more atherosclerosis parameters or perilesional tissue parameters, where the one or more atherosclerosis parameters include one or more of atherosclerosis presence, location, extent, severity, or type; and categorizing, by the computer system, the subject's baseline ASCVD risk into one or more predefined categories of ASCVD risk. determining, by the computer system, an initial personalized proposed treatment for the subject based at least in part on the subject's categorized baseline ASCVD risk, where the initial personalized proposed treatment for the subject includes one or more of a pharmacotherapy, a lifestyle therapy, or an interventional therapy; evaluating, by the computer system, the subject's response to the determined initial personalized proposed treatment by subsequent CCTA analysis using one or more quantitative imaging algorithms and comparing the subsequent CCTA analysis results with the baseline CCTA analysis results, where the subsequent CCTA analysis is performed after administering the determined initial personalized proposed treatment to the subject, where the subject response is evaluated based on one or more of ASCVD progression, stabilization, or resolution; evaluating and comparing; and based at least in part on the subject response evaluated by the computer system.determining a continued personalized suggested treatment for the subject, where when the assessed subject response includes progression of ASCVD, the continued personalized suggested treatment includes a higher tiered approach than the initial personalized suggested treatment, where the continued personalized suggested treatment includes one or more of a pharmacotherapy, a lifestyle therapy, or an interventional therapy, wherein the computer system includes a computer processor and an electronic storage medium;

[0017] In some embodiments, a computer-implemented method for determining volumetric stenosis severity in the presence of atherosclerosis based on non-invasive medical image analysis for risk assessment of coronary artery disease (CAD) for a subject using a normalization device, wherein the normalization device improves accuracy of the non-invasive medical image analysis. The method includes accessing, by a computer system, medical images of a coronary artery region of the subject, where the medical images of the coronary artery region of the subject are non-invasively acquired; identifying, by the computer system, one or more segments of a coronary artery and one or more regions of plaque within the medical images of the coronary artery region of the subject; and determining, by the computer system, for the identified one or more segments of the coronary artery, a luminal wall boundary in the presence of the one or more regions of plaque and a hypothetical normal arterial boundary if the one or more regions of plaque were not present, where the determined luminal wall boundary and the hypothetical normal arterial boundary include a taper of the one or more segments of the coronary artery, and the determined luminal wall boundary is a boundary of the one or more regions of plaque. and quantifying, by the computer system, a lumen volume for the identified one or more segments of the coronary artery based on the determined lumen wall boundary, where the quantified lumen volume takes into account tapering of the one or more segments of the coronary artery and boundaries of the one or more regions of plaque; quantifying, by the computer system, an assumed normal vessel volume for the identified one or more segments of the coronary artery based on the determined assumed normal arterial boundary, where the quantified assumed normal vessel volume takes into account tapering of the one or more segments of the coronary artery; determining, by the computer system, a volumetric stenosis by determining a percentage or ratio of the quantified lumen volume compared to the assumed normal vessel volume for the identified one or more segments of the coronary artery; and determining, by the computer system, a risk of CAD for the subject based at least in part on the determined volumetric stenosis for the identified one or more segments of the coronary artery.The computer system includes a computer processor and an electronic storage medium.

[0018] In some embodiments, a computer-implemented method for quantifying ischemia in a subject based on non-invasive medical image analysis using a normalization device, where the normalization device improves accuracy of the non-invasive medical image analysis. The method includes accessing, by a computer system, medical images of a coronary artery region of the subject, where the medical images of the coronary artery region of the subject are non-invasively acquired; identifying, by the computer system, one or more segments of the coronary artery and one or more regions of plaque within the medical images of the coronary artery region of the subject; and quantifying, by the computer system, a proximal volume of a proximal section and a distal volume of a distal section along the one or more segments of the coronary artery, where the proximal section is free of the one or more regions of plaque and the distal section is free of the one or more regions of plaque. the computer system comprising at least one of: accessing, by the computer system, an assumed velocity of blood flow in the proximal section; quantifying, by the computer system, a velocity of blood flow in the distal section based at least in part on the assumed velocity of blood flow in the proximal section along the one or more segments of the coronary artery, the quantified proximal volume of the proximal section, and the distal volume of the distal section; determining, by the computer system, a velocity time integral of the blood flow in the distal section based at least in part on the quantified velocity of blood flow in the distal section; and quantifying, by the computer system, ischemia along the one or more segments of the coronary artery based at least in part on the determined velocity time integral of the blood flow in the distal section, wherein the computer system comprises a computer processor and an electronic storage medium.

[0019] For purposes of this summary, certain aspects, advantages, and novel features of the invention are described herein. It should be understood that not all such advantages may necessarily be achieved in accordance with any particular embodiment of the invention. Thus, for example, one skilled in the art will recognize that the invention may be embodied or implemented in a manner that achieves one or more advantages taught herein, but not necessarily achieves other advantages that may be taught or presented herein.

[0020] All of these embodiments are intended to be within the scope of the invention disclosed herein. These and other embodiments will become readily apparent to those of ordinary skill in the art from the following detailed description, which refers to the accompanying drawings. The invention is not limited to any particular disclosed embodiment.

[0021] The disclosed aspects are described below in conjunction with the accompanying drawings, which are incorporated in and constitute a part of this specification, and are provided to illustrate and provide a further understanding of the example embodiments, but not to limit the disclosed aspects, in which like reference numerals represent like elements unless otherwise specified. [Brief description of the drawings]

[0022] [Figure 1] 1 is a flow chart illustrating an overview of an example embodiment of a method for medical image analysis, visualization, risk assessment, disease tracking, treatment generation, and / or patient report generation.

[0023] [Figure 2A] 1 is a flow chart illustrating an overview of an example embodiment of a method for analyzing and classifying plaque from medical images.

[0024] [Figure 2B] 1 is a flow chart illustrating an overview of one embodiment of a method for determining non-calcified plaque from non-contrast CT images.

[0025] [Figure 3A]1 is a flow chart illustrating an overview of an example embodiment of a method for risk assessment based on medical image analysis.

[0026] [Figure 3B] 1 is a flow chart illustrating an overview of one example embodiment of a method for quantifying atherosclerosis based on medical image analysis.

[0027] [Figure 3C] 1 is a flow chart illustrating an overview of one embodiment of a method for quantifying stenosis and generating a CAD-RADS score based on medical image analysis.

[0028] [Figure 3D] 1 is a flow chart illustrating an overview of an example embodiment of a method for disease tracking based on medical image analysis.

[0029] [Figure 3E] 1 is a flow chart illustrating an overview of an example embodiment of a method for determining the cause of a change in calcium score based on medical image analysis.

[0030] [Figure 4A] 1 is a flow chart illustrating an overview of an example embodiment of a method for prognosing a cardiovascular event based on medical image analysis.

[0031] [Figure 4B] 1 is a flow chart illustrating an overview of an example embodiment of a method for determining patient-specific stent parameters based on medical image analysis.

[0032] [Figure 5A] 1 is a flow chart illustrating an overview of an example embodiment of a method for generating a medical report regarding patient characteristics based on medical image analysis.

[0033] [Fig. 5B-5I] FIG. 1 illustrates an embodiment of an example of a medical report regarding patient characteristics generated based on medical image analysis.

[0034] [Figure 6A] FIG. 13 illustrates an example of a user interface that can be generated and displayed on the system and has multiple panels (images) that can show various corresponding views of a patient's arteries.

[0035] [Figure 6B] FIG. 13 illustrates an example of a user interface having multiple panels that can show various corresponding views of a patient's arteries that can be generated and displayed on the system.

[0036] [Figure 6C-6E] FIG. 2 shows certain details of the multiplanar reconstruction (MPR) vascular image of the second panel and certain functionality associated with this image.

[0037] [Figure 6F] FIG. 1 illustrates an example of a three-dimensional (3D) rendering of a coronary artery tree that allows a user to view the vessels and modify their levels.

[0038] [Figure 6G] FIG. 13 shows an example of a user interface panel that provides shortcut commands that a user may use while analyzing information in the user interface for axial, sagittal, and coronal views of the coronary arteries.

[0039] [Figure 6H] Figure 1 shows an example of a user interface panel for viewing DICOM images in three anatomical planes: axial, coronal, and sagittal.

[0040] [Figure 6I] FIG. 13 shows an example of a user interface panel showing a cross-sectional image of a vessel with a graphical overlay of extracted features of the vessel.

[0041] [Figure 6J] FIG. 13 illustrates an example of a toolbar that allows a user to select different vessels for inspection and analysis.

[0042] [Figure 6K] FIG. 6J shows an example of the series selection panel of the user interface in an expanded view of the toolbar shown in FIG. 6J, allowing the user to expand the menu and see all of the series (sets of images) available for a particular patient's examination and analysis.

[0043] [Figure 6L] FIG. 13 illustrates an example of a selection panel that may be displayed on a user interface that may be used to select vessel segments for analysis.

[0044] [Figure 6M] FIG. 13 shows an example of a panel that can be displayed on the user interface to add new vessels onto an image.

[0045] [Figure 6N] FIG. 11 shows an example of two panels that can be displayed on the user interface to name or re-name vessels in a 3-D arterial tree.

[0046] [Figure 7A] FIG. 13 illustrates an example of an editing toolbar that allows a user to modify and improve the accuracy of findings resulting from CT scans that are then processed by an analyst using machine learning algorithms.

[0047] [Figure 7B-7C] FIG. 13 illustrates an example of the specific functionality of the tracker tool.

[0048] [Figure 7D-7E] FIG. 13 illustrates the specific functionality of the vessel and lumen wall tools used to modify the lumen and vessel wall contours.

[0049] [Figure 7F] FIG. 14 illustrates the Vascular Snap tool button (left) and the Lumen Snap tool button (right) on the user interface that can be used to launch tools.

[0050] [Figure 7G] FIG. 13 illustrates an example of a panel that can be displayed on the user interface while using the Lumen Snap tool of the Vascular Snap tool.

[0051] [Figure 7H] FIG. 11 shows an example of a user interface panel that can be displayed while using the segment tool, allowing for marking the boundaries between individual coronary segments on the MPR.

[0052] [Figure 7I] FIG. 13 shows an example of a user interface panel that allows different names to be selected for the segments.

[0053] [Figure 7J] FIG. 13 shows an example of a user interface panel that can be displayed while using the stenosis tool, allowing the user to indicate a marker marking the extent of stenosis in a blood vessel.

[0054] [Figure 7K] FIG. 13 shows an example of a stenosis button on the user interface that can be used to drop five evenly spaced stenosis markers.

[0055] [Figure 7L] FIG. 13 shows an example of a stenosis button on the user interface that can be used to drop stenosis markers based on the user edited lumen and vessel wall contours.

[0056] [Figure 7M] FIG. 13 shows stenosis markers on a segment in a curved multiplanar vascular (CMPR) image.

[0057] [Figure 7N] FIG. 13 shows an example of a user interface panel that can be displayed while using the plaque overlay tool.

[0058] [Fig. 7O-7P] FIG. 14 illustrates the buttons on the user interface that can be selected for the plaque threshold.

[0059] [Figure 7Q] FIG. 13 illustrates a panel of a user interface that can receive user input to adjust plaque threshold levels for low-density plaque, non-calcified plaque, and calcified plaque.

[0060] [Figure 7R] FIG. 13 is a cross-sectional view of a blood vessel showing the extent of plaque displayed in the user interface according to a plaque threshold.

[0061] [Figure 7S] FIG. 13 shows a panel that can display the plaque threshold in a vessel statistics panel that contains information about the vessel being viewed.

[0062] [Figure 7T] FIG. 13 shows a panel showing a cross-sectional image of a blood vessel that can be displayed while using a centerline tool that allows adjustment of the center of the lumen.

[0063] [Fig. 7U-7W]FIG. 7U shows an example of an image that may be displayed when extending a vascular centerline; FIG. 7V shows an example of an image that may be displayed when saving or deleting centerline edits; and FIG. 7W shows an example of a CMPR image that may be displayed when editing a vascular centerline.

[0064] [Figure 7X] FIG. 13 shows an example of a panel that can be displayed while using a chronic total occlusion (CTO) tool used to show a portion of an artery that has 100% stenosis and no detectable blood flow.

[0065] [Figure 7Y] FIG. 13 shows an example of a panel that can be displayed while using the stent tool, allowing the user to mark the extent of the stent within the vessel.

[0066] [Fig. 7Z-7AA] FIG. 13 shows an example of a panel that can be displayed while using an exclusion tool, which allows excluding parts of a vessel from the analysis, for example due to image aberrations.

[0067] [Fig. 7AB-7AC] 7A-7C are diagrams showing examples of additional panels that may be displayed while using the exclusion tool; FIG. 7AB shows a panel that may be used to add a new exclusion, and FIG. 7AC shows a panel that may be used to add a reason for the exclusion.

[0068] [Fig. 7AD-7AG]7A-7C are diagrams showing examples of panels that may be displayed while using a distance tool that can be used to measure the distance between two points on an image, e.g., FIG. 7AD shows a distance tool used to measure distance on an SMPR image, FIG. 7AE shows a distance tool used to measure distance on a CMPR image, FIG. 7AF shows a distance tool used to measure distance on a cross-sectional view of a blood vessel, and FIG. 7AG shows a distance tool used to measure distance on an axial image.

[0069] [Figure 7AH] FIG. 13 shows the "Vascular Statistics" portion (button) of the panel which can be selected to display the Vascular Statistics tab.

[0070] [Figure 7AI] FIG. 13 illustrates the Vascular Statistics tab.

[0071] [Figure 7AJ] FIG. 13 illustrates functionality in the Vascular Statistics tab that allows the user to click through the details of multiple lesions.

[0072] [Figure 7AK] FIG. 13 further illustrates an example of a vessel panel that a user can use to toggle between vessels.

[0073] [Figure 8A] FIG. 1 shows an example of a panel of the user interface showing stenosis, atherosclerosis, and CAD-RADS results of the analysis.

[0074] [Figure 8B] FIG. 1 illustrates an example of a portion of a panel displayed on a user interface that allows selection of a region or combination of regions (e.g., left aorta (LM), left anterior descending artery (LAD), left circumflex artery (LCx), right coronary artery (RCA)) in accordance with various embodiments.

[0075] [Figure 8C] FIG. 13 shows an example of a panel that can be displayed on the user interface showing an animated representation of a coronary artery tree ("animated arterial tree").

[0076] [Figure 8D] FIG. 13 shows an example of a panel that can be displayed on the user interface, showing region selection using an animated arterial tree.

[0077] [Figure 8E] FIG. 13 illustrates an example panel that can be displayed on the user interface showing a region-by-region summary.

[0078] [Figure 8F] FIG. 1 shows an example panel that can be displayed on the user interface showing the SMPR image of a selected vessel, and the corresponding statistics for the selected vessel.

[0079] [Figure 8G] FIG. 13 shows an example of a portion of a panel that may be displayed in a user interface showing the presence of a stent, displayed at the segment level.

[0080] [Figure 8H] FIG. 13 illustrates an example of a portion of a panel that can be displayed in a user interface, showing the presence of a CTO at the segment level.

[0081] [Figure 8I] FIG. 13 shows an example of a portion of a panel that can be displayed in a user interface, indicating left or right dominance of a patient.

[0082] [Figure 8J] FIG. 13 shows an example of a panel that can be displayed on a user interface, showing an animated arterial tree indicating the anomalies found.

[0083] [Figure 8K] FIG. 8J shows an example of a portion of a panel that may be displayed on the panel of FIG. 8J that may be selected to show details of the anomaly.

[0084] [Figure 9A] FIG. 1 illustrates an example of an atherosclerosis panel that can be displayed on a user interface, displaying a summary of atherosclerosis information based on analysis.

[0085] [Figure 9B] FIG. 13 illustrates an example of a vessel selection panel that can be used to select vessels for which a summary of atherosclerosis information is displayed on a segment-by-segment basis.

[0086] [Figure 9C] FIG. 13 illustrates an example of a panel that can be displayed on a user interface showing atherosclerosis information per segment.

[0087] [Figure 9D] FIG. 13 shows an example of a panel that can be displayed on the user interface containing patient-specific data on stenosis.

[0088] [Figure 9E] FIG. 13 shows an example of a portion of a panel that may be displayed on a user interface where, once a count is selected (e.g., by hovering over the numbers), segment details are displayed.

[0089] [Figure 9F] FIG. 13 shows an example of a portion of a panel that can be displayed on a user interface, showing stenosis per segment in a graphical format, for example a bar graph of stenosis per segment.

[0090] [Figure 9G] FIG. 13 illustrates another example of a panel that can be displayed on the user interface, showing vascular information, such as percent diameter stenosis and minimum lumen diameter.

[0091] [Figure 9H] FIG. 13 shows an example of a portion of a panel that can be displayed on a user interface showing a legend of percent diameter stenosis.

[0092] [Figure 9I] FIG. 13 shows an example of a panel that can be displayed on the user interface showing minimum and reference luminal diameters.

[0093] [Figure 9J] FIG. 9I shows a portion of the panel shown in FIG. 9I and illustrates how details of a particular minimum luminal diameter can be quickly and efficiently displayed by selecting (e.g., by mouse-over) the desired graphic of the lumen.

[0094] [Figure 9K] FIG. 13 shows an example of a panel that can be displayed in a user interface showing CADS-RADS score selection.

[0095] [Figure 9L] FIG. 13 shows an example of a panel that can be displayed in the user interface showing further CAD-RADS details generated in the analysis.

[0096] [Figure 9M] FIG. 13 illustrates an example of a panel that can be displayed in the user interface, showing a table indicating the quantitative stenosis and vascular output determined during the analysis.

[0097] [Figure 9N] FIG. 13 shows an example of a panel that can be displayed in a user interface showing a table showing quantitative plaque output.

[0098] [Figure 10] 10 is a flow chart illustrating a process 1000 for analyzing and displaying CT images and corresponding information.

[0099] [Figure 11A-11B] 11A and 11B are example CT images illustrating how plaque can appear differently depending on the image acquisition parameters used to capture the CT images; FIG. 11A shows a CT image reconstructed using filtered back projection, and FIG. 11B shows the same CT image reconstructed using iterative reconstruction.

[0100] [Fig. 11C-11D] FIG. 11C shows another example illustrating that plaque can appear different in a CT image depending on the image acquisition parameters used to capture the CT image; FIG. 11C shows a CT image reconstructed using iterative reconstruction; and FIG. 11D shows the same image reconstructed using machine learning.

[0101] [Figure 12A] FIG. 2 is a block diagram representing one embodiment of a normalization device that can be configured to normalize medical images for use in the methods and systems described herein.

[0102] [Figure 12B] FIG. 1 is a perspective view illustrating an embodiment of a normalization device including a multi-layer substrate.

[0103] [Figure 12C] FIG. 12C is a cross-sectional view of the normalization device of FIG. 12B showing various compartments positioned therein to hold samples of known substances used during normalization.

[0104] [Figure 12D]FIG. 2 is a top view showing an example arrangement of multiple compartments within a normalization device, where in the illustrated embodiment the multiple compartments are arranged in a rectangular or grid pattern.

[0105] [Figure 12E] FIG. 13 is a top view showing another example arrangement of multiple compartments within a normalization device; in the illustrated embodiment, the multiple compartments are arranged in a circular pattern.

[0106] [Figure 12F] FIG. 13 is a cross-sectional view illustrating another embodiment of a normalization device exhibiting various features, including adjacently disposed compartments, a self-sealing, fillable compartment, and compartments of various sizes.

[0107] [Figure 12G] FIG. 13 is a perspective view showing one embodiment of an attachment mechanism of the normalization device, in which a hook-and-loop fastener is used to secure the substrate of the normalization device to the fastener of the normalization device.

[0108] [Fig. 12H-12I] FIG. 1 illustrates an embodiment of a normalization device including an indicator configured to indicate an expiration status of the normalization device.

[0109] [Figure 12J] 1 is a flowchart illustrating an example method for normalizing medical images for algorithm-based medical imaging analysis, where normalization of medical images improves accuracy of algorithm-based medical imaging analysis.

[0110] [Figure 13] FIG. 1 is a block diagram illustrating one embodiment of a system for medical image analysis, visualization, risk assessment, disease tracking, treatment generation, and / or patient report generation.

[0111] [Figure 14]FIG. 1 is a block diagram illustrating an embodiment of a computer hardware system configured to execute software to implement one or more embodiments of a system for medical image analysis, visualization, risk assessment, disease tracking, treatment generation, and / or patient report generation.

[0112] [Figure 15] FIG. 1 illustrates an embodiment of a normalization device.

[0113] [Figure 16] FIG. 1 is a system diagram illustrating various components of an example system for automatically generating patient medical reports, such as patient medical reports based on CT scans and analysis, utilizing certain systems and methods described herein.

[0114] [Figure 17] FIG. 2 is a block diagram illustrating an example of data flow functionality for generating a patient medical report based on one or more scans of a patient, patient information, physician scan analysis, and / or previous test results.

[0115] [Figure 18A] 3 is a block diagram of a first portion of a process for generating a medical report using the functions and data described with reference to FIG. 2, according to some embodiments.

[0116] [Figure 18B] 3 is a block diagram of a second portion of a process for generating a medical report using the functions and data described with reference to FIG. 2 according to some embodiments.

[0117] [Figure 18C] 3 is a block diagram of a third portion of a process for generating a medical report using the functions and data described with reference to FIG. 2 according to some embodiments.

[0118] [Figure 18D] FIG. 1 illustrates various parts that can make up a medical report and inputs can be provided by a physician and by patient information or input.

[0119] [Figure 18E] FIG. 1 is a schematic diagram illustrating an example of medical report generation data flow and data communications used to generate a report.

[0120] [Figure 18F] FIG. 1 illustrates multiple structures for storing information used in a medical report, the information being associated with a patient based on one or more characteristics of the patient, the patient's medical condition, and / or input from the patient or physician.

[0121] [Figure 19A] FIG. 1 illustrates an example process for determining risk assessment using sequential imaging of non-contiguous arterial beds of a patient, according to some embodiments.

[0122] [Figure 19B] FIG. 1 shows an example in which sequential non-contiguous arterial bed imaging is performed on the coronary arteries.

[0123] [Figure 19C] FIG. 13 is an example diagram of a process for determining risk assessment using sequential imaging of non-contiguous arterial beds, according to some embodiments.

[0124] [Figure 19D] FIG. 13 is an example diagram of a process for determining risk assessment using sequential imaging of non-contiguous arterial beds, according to some embodiments.

[0125] [Figure 19E]FIG. 1 is a block diagram depicting an embodiment of a computer hardware system configured to execute software for implementing one or more embodiments of systems and methods for determining risk assessment using sequential imaging of non-contiguous arterial beds of a patient.

[0126] [Figure 20A] FIG. 1 illustrates one or more characteristics of an exemplary ischemic pathway.

[0127] [Figure 20B] FIG. 1 is a block diagram depicting one or more triggers and one or more time sequences of an ischemic outcome utilized by exemplary embodiments described herein.

[0128] [Figure 20C] 1 is a block diagram depicting one or more features of an example embodiment for determining ischemia by separately weighting different factors.

[0129] [Figure 20D] FIG. 1 is a block diagram depicting one or more features of an example embodiment for calculating a global ischemic index.

[0130] [Figure 20E] 1 is a flow chart illustrating an overview of an exemplary embodiment of a method for generating a global ischemia index for a subject and using the index to assist in assessing the subject's risk of ischemia.

[0131] [Figure 21] 1 is a flow chart illustrating an overview of an exemplary embodiment of a method for generating a coronary artery disease (CAD) score for a subject and using the score to assist in assessing the subject's risk of CAD.

[0132] [Figure 22A]FIG. 1 illustrates an example of tracking plaque decay for analysis and / or treatment of coronary and / or other vascular disease.

[0133] [Figure 22B] 1 is a flow chart illustrating an overview of an exemplary embodiment of a method for processing images.

[0134] [Figure 23A] FIG. 1 illustrates an exemplary embodiment of a system and method for determining a treatment for reducing cardiovascular risk and / or events.

[0135] [Fig. 23B-23C] FIG. 1 illustrates an exemplary embodiment of atherosclerosis severity definitions or categories used by an exemplary embodiment of a system and method for determining treatments to reduce cardiovascular risk and / or events.

[0136] [Figure 23D] FIG. 1 illustrates an exemplary embodiment of definitions or categories of disease progression, stabilization, and / or disappearance used by an exemplary embodiment of a system and method for determining treatments for reducing cardiovascular risk and / or events.

[0137] [Figure 23E] FIG. 1 illustrates an exemplary embodiment of a time to treatment target for an exemplary embodiment of a system and method for determining a treatment for reducing cardiovascular risk and / or events.

[0138] [Fig. 23F-23G] FIG. 1 illustrates an exemplary embodiment of a treatment employing lipid-lowering drugs and / or treatments generated by an exemplary embodiment of a system and method for determining a treatment for reducing cardiovascular risk and / or events.

[0139] [Fig. 23H-23I]FIG. 1 illustrates an exemplary embodiment of a treatment employing diabetes drugs and / or treatments generated by an exemplary embodiment of a system and method for determining a treatment for reducing cardiovascular risk and / or events.

[0140] [Figure 23J] 1 is a flow chart showing an overview of an exemplary embodiment of a method for determining a treatment for reducing cardiovascular risk and / or events.

[0141] [Figure 24A] FIG. 1 is a schematic diagram of an artery.

[0142] [Figure 24B] FIG. 1 illustrates an embodiment of determining percentage stenosis and remodeling index.

[0143] [Figure 24C] FIG. 1 is a schematic diagram of an artery.

[0144] [Figure 24D] FIG. 1 is a schematic diagram showing an artery with a long atherosclerotic region of plaque.

[0145] [Figure 24E] FIG. 11 is an example illustrating how an inaccurately estimated R0 can significantly affect the resulting percent stenosis and / or remodeling index.

[0146] [Figure 24F] FIG. 1 is a schematic diagram showing lumen diameter versus outer wall diameter.

[0147] [Figure 24G] FIG. 1 is a schematic diagram showing the calculation of an estimated reference diameter along a vessel in which plaque is present.

[0148] [Fig. 24H] FIG. 1 is a schematic diagram illustrating an embodiment of determining volumetric stenosis.

[0149] [Figure 24I] FIG. 1 is a schematic diagram illustrating an embodiment of determining volumetric stenosis.

[0150] [Figure 24J] FIG. 1 is a schematic diagram illustrating an embodiment of determining volumetric remodeling.

[0151] [Figure 24K] FIG. 1 illustrates an embodiment of coronary blood volume assessment based on total coronary artery volume.

[0152] [Figure 24L] FIG. 1 illustrates an embodiment of regional or artery specific volumetric based coronary blood volume assessment.

[0153] [Figure 24M] FIG. 1 illustrates an embodiment of coronary blood volume assessment based on intra-arterial % fractional blood volume.

[0154] [Figure 24N] FIG. 1 illustrates an embodiment of an assessment of coronary blood volume.

[0155] [Fig. 24O] FIG. 1 illustrates an embodiment of the assessment of % vascular volume stenosis as a measure of ischemia.

[0156] [Figure 24P] FIG. 1 illustrates an embodiment of an assessment of pressure difference across a lesion as a measure of ischemia.

[0157] [Figure 24Q] FIG. 1 illustrates an embodiment of the application of the continuity equation to a coronary artery.

[0158] [Figure 24R] 1 is a flow chart illustrating an overview of one embodiment of a method for determining volumetric stenosis and / or volumetric vascular remodeling.

[0159] [Figure 24S] 1 is a flow chart illustrating an overview of one embodiment of a method for determining ischemia.

[0160] [Figure 25A] 1 is a flow chart showing a process for determining a risk index that atherosclerotic lesions will contribute to myocardial infarction or other major adverse cardiovascular events.

[0161] [Figure 25B] FIG. 1 is a schematic diagram of the human heart showing certain coronary arteries.

[0162] [Figure 25C] 1 is a flow chart showing a process for determining a myocardial risk index posed by atherosclerotic lesions.

[0163] [Figure 25D] 1 is a flow chart showing a process for determining the myocardial risk index of a segment contributed by atherosclerotic lesions.

[0164] [Figure 25E] 1 is a flow chart illustrating a process for determining the risk of an adverse clinical event caused by atherosclerotic lesions.

[0165] [Figure 25F] 1 is a flow chart illustrating a process for updating the risk of adverse clinical events caused by atherosclerotic lesions.

[0166] [Figure 25G]FIG. 1 is a block diagram depicting an embodiment of a computer hardware system configured to execute software for implementing one or more embodiments of systems, devices, and methods for determining myocardial risk factors from image-based quantification and characterization of coronary artery atherosclerosis, vascular morphology, and myocardium.

[0167] [Figure 26] 1 is a flowchart illustrating a process for analyzing CFD-based indications of ischemia using characterization of atherosclerosis and vascular morphology.

[0168] [Figure 27A] 1A-1D are block diagrams illustrating exemplary embodiments of systems, devices, and methods for determining patient-specific and / or subject-specific coronary artery disease (CAD) risk factor targets from image-based quantified atherosclerosis phenotyping.

[0169] [Figure 27B] FIG. 27A is a block diagram of an example computing system that can be used to implement the systems, processes, and methods described herein with respect to the functionality described with reference to FIG.

[0170] [Fig. 28A-28B] FIG. 1 illustrates an exemplary embodiment of identification of coronary artery and aortic disease / atherosclerosis identified on a coronary CT angiogram (CCTA) utilizing embodiments of the systems, devices, and methods described herein.

[0171] [Figure 28C] 1 is a flow chart illustrating exemplary embodiments of systems, devices, and methods for image-based diagnosis, risk assessment, and / or characterization of major adverse cardiovascular events.

[0172] [Figure 28D]1 is a flow chart illustrating exemplary embodiments of systems, devices, and methods for image-based diagnosis, risk assessment, and / or characterization of major adverse cardiovascular events.

[0173] [Figure 28E] 1 is a flow chart illustrating exemplary embodiments of systems, devices, and methods for image-based diagnosis, risk assessment, and / or characterization of major adverse cardiovascular events.

[0174] [Figure 28F] FIG. 1 is a block diagram depicting an embodiment of a computer hardware system configured to execute software for implementing one or more embodiments of the systems, devices, and methods described herein.

[0175] [Figure 29A] 1 is a block diagram illustrating an example embodiment of a system, device, and method for improving the accuracy of CAD measurements in non-invasive imaging.

[0176] [Figure 29B] FIG. 1 is a block diagram depicting an embodiment of a computer hardware system configured to execute software for improving accuracy of CAD measurements in non-invasive imaging.

[0177] [Figure 30A] 1A-1D are block diagrams illustrating exemplary embodiments of systems, devices, and methods for longitudinal image-based phenotyping to enhance drug discovery or development.

[0178] [Figure 30B]FIG. 1 is a block diagram depicting an embodiment of a computer hardware system configured to execute software for implementing one or more embodiments of systems, devices, and methods for determining patient-specific coronary artery disease (CAD) risk factor targets from image-based quantification and characterization of coronary atherosclerosis burden, type, and / or progression rate. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0179] Although several embodiments, examples, and exemplifications are disclosed below, those skilled in the art will understand that the invention described herein extends beyond the specifically disclosed embodiments, examples, and exemplifications, and includes other uses of the invention, as well as obvious modifications and equivalents of the invention. The embodiments of the invention are described with reference to the accompanying drawings, in which like numerals refer to like elements throughout the drawings. The terms used in the description presented herein should not be construed in any restrictive or limiting manner simply because they are used in conjunction with the detailed description of certain specific embodiments of the invention. In addition, the embodiments of the invention may comprise several novel features, any feature being solely responsible for desirable attributes or being essential to practice the invention described herein.

[0180] Introduction Disclosed herein are systems, methods, and devices for medical image analysis, diagnosis, severity classification, decision making, and / or disease tracking. Coronary heart disease affects over 17.6 million Americans. The current trend in treating cardiovascular health problems is nearly two-fold. First, physicians typically examine a patient's cardiovascular health at a macro level, for example, by analyzing biochemistry or blood content or biomarkers to determine whether there are high levels of cholesterol elements in the patient's bloodstream. In response to high levels of cholesterol, some physicians prescribe one or more drugs, such as statins, as part of a treatment plan to reduce what are perceived to be high levels of cholesterol elements in the patient's bloodstream.

[0181] A second general trend for currently treating cardiovascular health problems involves physicians assessing a patient's cardiovascular health by using angiograms to identify large blockages in the patient's various arteries. In response to finding large blockages in various arteries, the physician sometimes performs an angioplasty procedure in which a balloon catheter is guided to the point where the blood vessel is narrowed. Once properly positioned, the balloon is inflated to compress or flatten plaque or fatty material into the artery wall and / or stretch the artery open to increase blood flow through the vessel and / or to the heart. Sometimes, the balloon is used to position and expand a stent within the vessel to compress plaque and / or keep the vessel open to allow more blood to flow. Approximately 500,000 cardiac stent procedures are performed in the United States each year.

[0182] However, a recent $100 million federally funded study questions whether current trends in cardiovascular disease treatment are the most effective treatment for all types of patients. The recent study involved more than 5,000 patients with moderate to severe stable heart disease from 320 locations in 37 countries, and provided new evidence that stent and bypass surgery procedures are unlikely to be more effective than drugs combined with lifestyle changes in patients with stable heart disease. Therefore, patients with stable heart disease may be better off forgoing invasive surgical procedures such as angioplasty and / or cardiac bypass, and instead receiving prescriptions for heart medications such as statins and making certain lifestyle changes, such as regular exercise. This new treatment plan could impact thousands of patients worldwide. It is estimated that one-fifth of the estimated 500,000 cardiac stent procedures performed annually in the United States are for patients with stable heart disease. Furthermore, it is estimated that 25% of the estimated 100,000 patients with stable heart disease, approximately 23,000 people, do not experience chest pain. Thus, over 20,000 patients annually may forego invasive surgical procedures or may avoid complications as a result of such procedures.

[0183] It may be important to better understand a patient's cardiovascular disease to determine whether the patient should forgo invasive surgical procedures and instead opt for a drug treatment plan and / or develop a more effective treatment plan. In particular, it may be advantageous to better understand the health of a patient's arterial blood vessels. For example, it is helpful to understand whether the plaque buildup in a patient's body is mostly fatty buildup or mostly calcified material buildup, because the former situation requires treatment with cardiac drugs such as statins, whereas the latter situation should further monitor the patient periodically without prescribing cardiac drugs or implanting any stents. However, if the plaque buildup is significant enough to cause severe stenosis or narrowing of the arterial blood vessels, which may block blood flow to the heart muscle, an invasive angioplasty procedure to implant a stent may be necessary because such patients may experience a heart attack or sudden cardiac death (SCD) if a stent is not implanted to enlarge the opening of the blood vessels. Sudden cardiac death is one of the leading causes of natural deaths in the United States, responsible for approximately 325,000 adult deaths annually, accounting for approximately half of all deaths from cardiovascular disease. Men are twice as likely to develop SCD than women. SCD typically occurs in people in their mid-30s to mid-40s. More than 50% experience sudden cardiac arrest without warning.

[0184] With millions of people suffering from heart disease, there is a need to better understand the overall health of arterial vessels in a patient's body, rather than simply knowing the blood chemistry or content of the blood flowing through such vessels. For example, in some embodiments of the systems, devices, and methods disclosed herein, arteries with "benign" or stable plaques or plaques containing hardened calcified content are considered not threatening to the patient's life, whereas "malignant" or unstable plaques or plaques containing fatty material are considered more dangerous to life, as such malignant plaques may rupture within the artery, thereby releasing such fatty material into the artery. When such fatty material is released into the bloodstream, it may cause inflammation, which may lead to blood clots. Blood clots within the artery may prevent blood from moving to the heart muscle, thereby causing a heart attack or other cardiac event. Furthermore, in some cases, it is generally more difficult for blood to flow through fatty plaques that have accumulated than it is for blood to flow through calcified plaques that have accumulated. Thus, there is a need to better understand and analyze the walls of a patient's arterial vessels.

[0185] Furthermore, while blood tests and drug treatment plans can help reduce cardiovascular health problems and mitigate cardiovascular events (e.g., heart attacks), such treatment methodologies are not perfect or complete in that they may misidentify and / or fail to locate or diagnose significant cardiovascular risk areas. For example, simply analyzing a patient's blood chemistry may not identify the patient as having an arterial vessel with a significant amount of malignant plaque of fatty deposit material along the vessel wall. Similarly, an angiogram may help identify the extent of stenosis or vascular narrowing, but may not clearly identify the area of ​​the arterial vessel wall where malignant plaque has significantly accumulated. The extent of such malignant plaque accumulation within the arterial vessel wall may be indicative of a patient at high risk of suffering a cardiovascular event, such as a heart attack. In certain circumstances, areas where malignant plaque is present may lead to rupture, releasing fatty material into the arterial bloodstream, which may in turn cause a blood clot within the artery. A blood clot within the artery may stop blood flow to the heart tissue, which may result in a heart attack. Thus, there is a need for new techniques to analyze arterial vessel walls and / or identify areas within the arterial vessel walls that contain plaque buildup, whether malignant or not.

[0186] Various systems, methods, and devices disclosed herein are directed to embodiments that address the above-mentioned challenges. In particular, various embodiments described herein relate to systems, methods, and devices for medical image analysis, diagnosis, severity classification, decision making, and / or disease tracking. In some embodiments, the systems, devices, and methods described herein are configured to identify one or more coronary arteries and / or plaque therein utilizing non-invasive medical imaging techniques, such as, for example, CT images, which may be input into a computer system configured to automatically and / or dynamically analyze the medical images. For example, in some embodiments, the system may be configured to automatically and / or dynamically analyze the medical images to identify, quantify, and / or classify the one or more coronary arteries and / or plaque utilizing one or more machine learning and / or artificial intelligence algorithms. In some embodiments, the system may be further configured to utilize the identified, quantified, and / or classified one or more coronary arteries and / or plaque, for example, using one or more artificial intelligence and / or machine learning algorithms, to generate a treatment plan, track disease progression, and / or provide medical reports regarding patient characteristics. In some embodiments, the system may be further configured to dynamically and / or automatically generate a visualization of the identified, quantified, and / or classified coronary artery(s) and / or plaque, e.g., in the form of a graphical user interface.Further, in some embodiments, the system may be configured to utilize a normalization device comprising one or more sections of one or more substances to calibrate medical images obtained from different medical imaging scanners and / or different scanning parameters or environments.

[0187] As discussed in further detail, the systems, devices, and methods described herein enable automated and / or dynamic quantified analysis of various parameters related to plaque, cardiovascular arteries, and / or other structures. More specifically, in some embodiments described herein, medical images of a patient, such as coronary CT images, can be taken at a medical institution. Rather than being viewed by a physician or performing a general evaluation of the patient, the medical images are transmitted to a back-end main server, which in some embodiments is configured to perform one or more analyses thereof in a reproducible manner. Thus, in some embodiments, the systems, methods, and devices described herein can provide quantified measurements of one or more features of the coronary CT images using an automated and / or dynamic process. For example, in some embodiments, the main server system can be configured to identify one or more blood vessels, plaque, and / or fat from the medical images. Based on the identified features, in some embodiments, the system can be configured to generate one or more quantified measurements from the raw medical images, such as, for example, the radiodensity of one or more regions of plaque, the identification of stable and / or unstable plaque, its volume, its surface area, its geometry, its heterogeneity, and / or others. In some embodiments, the system can also generate one or more quantified measurements of blood vessels from the raw medical images, such as, for example, diameter, volume, morphology, and / or others. Based on the identified features and / or quantified measurements, in some embodiments, the system can be configured to generate a risk assessment and / or track the progression of a plaque-based disease or condition, such as, for example, atherosclerosis, stenosis, and / or ischemia, using the raw medical images. Furthermore, in some embodiments, the system can be configured to generate a GUI visualization of one or more identified features and / or quantified measurements, such as a quantized color mapping of different features.In some embodiments, the systems, devices, and methods described herein are configured to utilize medical image-based processing to assess a subject's risk for cardiovascular events, major adverse cardiovascular events (MACE), rapid plaque progression, and / or non-response to drugs. In particular, in some embodiments, the system can be configured to automatically and / or dynamically assess such health risks of a subject by analyzing only non-invasively obtained medical images. In some embodiments, one or more of the processes can be automated using AI and / or ML algorithms. In some embodiments, one or more of the processes described herein can be performed reproducibly and within minutes. This is in stark contrast to today's existing standards, which do not produce reproducible prognoses or assessments, take significant amounts of time, and / or require invasive procedures.

[0188] Thus, in some embodiments, the systems, devices, and methods described herein can provide physicians and / or patients with specific quantified and / or measured data regarding a patient's plaque that does not exist today. For example, in some embodiments, the system can use, for example, radiodensity values ​​of pixels and / or regions within a medical image to provide specific numerical values ​​regarding stable and / or unstable plaque volume, its ratio to total vessel volume, percentage of stenosis, and / or the like. In some embodiments, such a detailed level of quantified plaque parameters resulting from image processing and downstream analysis can provide more accurate and useful tools to assess a patient's health and / or risk in entirely novel ways.

[0189] Overall Overview In some embodiments, the systems, devices, and methods described herein are configured to automatically and / or dynamically perform medical image analysis, diagnosis, severity classification, decision making, and / or disease tracking. Figure 1 is a flow chart illustrating an overview of an example embodiment of a method for medical image analysis, visualization, risk assessment, disease tracking, treatment generation, and / or patient report generation. As shown in Figure 1, in some embodiments, the system is configured to evaluate and / or analyze one or more medical images of a subject, such as, for example, medical images of a coronary artery region of the subject or patient.

[0190] In some embodiments, prior to obtaining the medical images, at block 102, a normalization device is attached to the subject and / or placed within the field of view of the medical imaging scanner. For example, in some embodiments, the normalization device can comprise one or more compartments containing one or more substances, such as water, calcium, and / or others. Additional details regarding normalization devices are provided below. A medical imaging scanner may create images having different scalable radiodensities for the same object. This may vary, for example, depending on the type of medical imaging scanner or device used, as well as the scanning parameters and / or environment on the particular day and / or time the scan was taken. As a result, even when two different scans of the same subject are taken, the resulting medical images may differ in brightness and / or darkness, which may compromise the accuracy of any analysis processed from the images. To account for such differences, in some embodiments, a normalization device comprising one or more known elements is scanned with the subject, and the resulting image of the one or more known elements may be used as a reference for translating, transforming, and / or normalizing the resulting image. As such, in some embodiments, the normalization device is attached to a subject and / or placed within the field of view of a medical imaging scan in a medical institution.

[0191] In some embodiments, at block 104, the medical institution then obtains one or more medical images of the subject. For example, the medical images can be of the coronary artery region of the subject or patient. In some embodiments, the systems disclosed herein can be configured to capture CT data from an image domain or a projection domain, such as, but not limited to, X-ray, dual energy computed tomography (DECT), spectral CT, photon counting detector CT, ultrasound such as echocardiography or intravascular ultrasound (IVUS), magnetic resonance (MR) imaging, optical coherence tomography (OCT), nuclear medicine imaging including positron emission tomography (PET) and single photon emission computed tomography (SPECT), near-field infrared spectroscopy (NIRS), and / or others. As used herein, the terms CT image data or CT scan data can be interchanged with any of the above-mentioned medical scanning modalities and processes that process such data through an artificial intelligence (AI) algorithmic system to generate processed CT image data. In some embodiments, data from these imaging modalities may include image domain data, projection domain data, and / or a combination of both, allowing cardiovascular phenotypes to be determined.

[0192] In some embodiments, at block 106, the medical institution may also obtain non-imaging data from the subject. For example, this may include blood tests, biomarkers, panomics, and / or others. In some embodiments, at block 108, the medical institution may transmit one or more medical images and / or other non-imaging data at block 108 to the main server system. In some embodiments, the main server system may be configured to receive and / or otherwise access the medical images and / or other non-imaging data at block 110.

[0193] In some embodiments, at block 112, the system may be configured to automatically and / or dynamically analyze one or more medical images that may be stored and / or accessed from the medical image database 100. For example, in some embodiments, the system may be configured to take raw CT image data and apply artificial intelligence (AI), machine learning (ML), and / or other physics-based algorithms to the raw CT data to identify, measure, and / or analyze various aspects of the arteries identified in the CT data. In some embodiments, inputting the raw medical image data involves uploading the raw medical image data to a cloud-based data repository system. In some embodiments, processing the medical image data involves using AI and / or ML algorithms to process the data in the cloud-based computing system. In some embodiments, the system can be configured to analyze the raw CT data within about 1 minute, about 2 minutes, about 3 minutes, about 4 minutes, about 5 minutes, about 6 minutes, about 7 minutes, about 8 minutes, about 0 minutes, about 10 minutes, about 15 minutes, about 20 minutes, about 30 minutes, about 35 minutes, about 40 minutes, about 45 minutes, about 50 minutes, about 55 minutes, about 60 minutes, and / or within a range defined by two of the above values.

[0194] In some embodiments, the system may be configured to utilize a vessel identification algorithm to identify and / or analyze one or more vessels in a medical image. In some embodiments, the system may be configured to utilize a coronary artery identification algorithm to identify and / or analyze one or more coronary arteries in a medical image. In some embodiments, the system may be configured to utilize a plaque identification algorithm to identify and / or analyze one or more plaque regions in a medical image. In some embodiments, the vessel identification algorithm, the coronary artery identification algorithm, and / or the plaque identification algorithm include AI and / or ML algorithms. For example, in some embodiments, the vessel identification algorithm, the coronary artery identification algorithm, and / or the plaque identification algorithm may be trained on a plurality of medical images in which one or more vessels, coronary arteries, and / or plaque regions are pre-identified. Based on such training, for example, in some embodiments by using a convolutional neural network, the system may be configured to automatically and / or dynamically identify the presence and / or parameters of vessels, coronary arteries, and / or plaque from raw medical images.

[0195] As such, in some embodiments, processing the medical images or raw CT scan data can include analyzing the medical images or CT data to determine and / or identify the presence and / or absence of particular arterial blood vessels within a patient's body. As a naturally occurring phenomenon, certain arteries may be present in a particular patient's body, while such particular arteries may not be present in other patients' bodies.

[0196] In some embodiments, at block 112, the system may be further configured to analyze the identified vessels, coronary arteries, and / or plaques, e.g., using AI and / or ML algorithms. In particular, in some embodiments, the system may be configured to determine one or more vessel morphology parameters, e.g., arterial remodeling, curvature, volume, width, diameter, length, and / or others. In some embodiments, the system may be configured to determine one or more plaque parameters, e.g., volume, surface area, geometry, radiodensity, volume-to-surface area ratio or function, heterogeneity index, and / or others, of one or more plaque regions shown in the medical image. "Radiodensity," as used herein, is a broad term referring to the relative inability of electromagnetic relations (e.g., x-rays) to pass through matter. With reference to an image, radiodensity values ​​refer to values ​​indicative of density of image data (e.g., film, print, or electronic format), and radiodensity values ​​of an image correspond to the density of material shown in the image.

[0197] In some embodiments, at block 114, the system may be configured to utilize the identified and / or analyzed vessels, coronary arteries, and / or plaque from the medical images to perform a time point analysis of the subject. In some embodiments, the system may be configured to use automatic and / or dynamic image processing of one or more medical images taken from a time point to identify and / or analyze one or more vessels, coronary arteries, and / or plaque and derive one or more parameters and / or classifications thereof. For example, as described in more detail herein, in some embodiments, the system may be configured to generate one or more quantification metrics of the plaque and / or classify the identified plaque regions as benign or malignant plaque. Further, in some embodiments, at block 114, the system may be configured to generate one or more treatment plans for the subject based on the analysis results. In some embodiments, the system may be configured to utilize one or more AI and / or ML algorithms to identify and / or analyze the vessels or plaque, derive one or more quantification metrics and / or classifications, and / or generate a treatment plan.

[0198] In some embodiments, if previous scans or medical images of the subject exist, the system may be configured to perform one or more time-based analyses, such as disease tracking, at block 126. For example, in some embodiments, if the system has access to one or more quantified parameters or classifications derived from previous scans or medical images of the subject, the system may be configured to compare them with one or more quantified parameters or classifications derived from the current scan or medical images to determine the progression and / or status of the subject's disease.

[0199] In some embodiments, at block 116, the system is configured to automatically and / or dynamically generate a graphical user interface (GUI) or other visualization of the analysis results, which may include, for example, identified vessels, plaque areas, coronary arteries, quantified metrics or parameters, risk assessment, proposed treatment plan, and / or any other analysis results discussed herein. In some embodiments, the system is configured to analyze the arteries present in the CT scan data and display various views of the arteries present in the patient's body, for example, in 10-15 minutes or less. In contrast, by way of example, visual evaluation of a CT scan to identify only stenosis without considering benign or malignant plaque or any other factors may take from 15 minutes to over an hour depending on skill level and may have significant variability across radiologists and / or cardiac imaging devices.

[0200] In some embodiments, the system may be configured to transmit the generated GUI or other visualization, analysis results, and / or treatment to a medical institution, block 118. In some embodiments, a physician at the medical institution may then review and / or confirm and / or correct the generated GUI or other visualization, analysis results, and / or treatment, block 120.

[0201] In some embodiments, the system may be configured to further generate and transmit a medical report of the patient characteristics to the patient at block 122, which the patient may receive at block 124. In some embodiments, the medical report of the patient characteristics may be dynamically generated based on analysis results derived from the medical image processing and analysis and / or other generated therefrom. For example, the patient-specific report may include identified vessels, plaque areas, coronary arteries, quantified metrics or parameters, risk assessment, proposed treatment plan, and / or any other analysis results discussed herein.

[0202] In some embodiments, one or more of the processes depicted in FIG. 1 may be repeated, for example, on the same patient at different times to track the progression and / or condition of the patient's disease.

[0203] Image processing-based classification of benign versus malignant plaques As mentioned above, in some embodiments, the systems, methods, and devices described herein are configured to automatically and / or dynamically identify and / or classify benign versus malignant or stable versus unstable plaques based on medical image analysis and / or processing. For example, in some embodiments, the system can be configured to utilize AI and / or ML algorithms to identify areas within an artery that exhibit plaque buildup within, along, within, and / or outside of the artery. In some embodiments, the system can be configured to identify the contour or boundary of the plaque buildup associated with the arterial vessel wall. In some embodiments, the system can be configured to draw or generate lines that define the shape and configuration of the plaque buildup associated with the artery. In some embodiments, the system can be configured to identify whether the plaque buildup is a particular type of plaque and / or the composition or characteristics of the particular plaque buildup. In some embodiments, the system can be configured to characterize the plaque binary, sequential, and / or sequentially. In some embodiments, the system may be configured to determine that the type of plaque accumulation identified is a "malignant" plaque by the nature of the dark or dark gray scale of the image corresponding to the plaque area and / or by determining its attenuation density (e.g., using the Hounsfield Units scale or otherwise). For example, in some embodiments, the system may be configured to identify a particular plaque as a "malignant" plaque if the brightness of the plaque is darker than a predetermined level. In some embodiments, the system may be configured to identify a benign plaque area based on the white shade and / or light gray scale nature of the area corresponding to the plaque accumulation. For example, in some embodiments, the system may be configured to identify a particular plaque as a "benign" plaque if the brightness of the plaque is lighter than a predetermined level.In some embodiments, the system may be configured to determine that dark areas of the CT scan relate to "malignant" plaque, while the system may be configured to identify areas of benign plaque that correspond to white areas. In some embodiments, the system may be configured to identify and determine the total area and / or volume of the identified total plaque, benign plaque, and / or malignant plaque in an arterial vessel or vessels. In some embodiments, the system may be configured to determine the length of the identified total plaque, benign plaque, and / or malignant plaque. In some embodiments, the system may be configured to determine the width of the identified total plaque, benign plaque, and / or malignant plaque. "Benign" plaque may be considered as such because it is less likely to cause a heart attack, less likely to show significant plaque progression, and / or less likely to be ischemic, among other things. Conversely, "malignant" plaque is considered as such because it is more likely to cause a heart attack, more likely to show significant plaque progression, and / or more likely to be ischemic, among other things. In some embodiments, a "benign" plaque may be considered as such because it is less likely to result in a reflow phenomenon upon coronary revascularization. Conversely, a "malignant" plaque may be considered as such because it is more likely to not result in a reflow phenomenon upon coronary revascularization.

[0204] FIG. 2A is a flow chart showing an overview of an example embodiment of a method for analyzing and classifying plaque from medical images, which may be obtained non-invasively. As shown in FIG. 2A, at block 202, in some embodiments, the system may be configured to access medical images, which may include a coronary artery region of a subject and / or may be stored in a medical image database 100. The medical image database 100 may be locally accessible by the system and / or may be remotely located and accessible through a network connection. The medical images may include images obtained using one or more modalities, such as, for example, CT, dual energy computed tomography (DECT), spectral CT, photon counting CT, x-ray, ultrasound, echocardiography, intravascular ultrasound (IVUS), magnetic resonance (MR) imaging, optical coherence tomography (OCT), nuclear medicine imaging, positron emission tomography (PET), single photon emission computed tomography (SPECT), or near-field infrared spectroscopy (NIRS). In some embodiments, the medical images include one or more of contrast-enhanced CT images, non-contrast CT images, MR images, and / or images obtained using any of the modalities mentioned above.

[0205] In some embodiments, the system may be configured to automatically and / or dynamically perform one or more analyses of the medical images as discussed herein. For example, in some embodiments, at block 204, the system may be configured to identify one or more arteries. The one or more arteries may include a coronary artery, a carotid artery, an aorta, a renal artery, a lower limb artery, an upper limb artery, and / or a cerebral artery, among others. In some embodiments, the system may be configured to utilize one or more AI and / or ML algorithms to automatically and / or dynamically identify one or more arteries or coronary arteries using image processing. For example, in some embodiments, the one or more AI and / or ML algorithms may be trained using a convolutional neural network (CNN) on a set of medical images in which the arteries or coronary arteries have been identified, thereby enabling the AI ​​and / or ML algorithm to automatically identify the arteries or coronary arteries directly from the medical images. In some embodiments, the arteries or coronary arteries are identified by size and / or location.

[0206] In some embodiments, at block 206, the system may be configured to identify one or more plaque regions in the medical images. In some embodiments, the system may be configured to utilize one or more AI and / or ML algorithms to automatically and / or dynamically identify one or more plaque regions using image processing. For example, in some embodiments, one or more AI and / or ML algorithms may be trained using a convolutional neural network (CNN) on a set of medical images in which plaque regions have been identified, thereby enabling the AI ​​and / or ML algorithms to automatically identify plaque regions directly from the medical images. In some embodiments, the system may be configured to identify a vessel wall and a lumen wall for each coronary artery identified in the medical images. In some embodiments, the system is then configured to determine the volume between the vessel wall and the lumen wall as plaque. In some embodiments, the system may be configured to identify plaque regions based on radiodensity values ​​commonly associated with plaque, for example, by setting a predetermined threshold or range of radiodensity values ​​commonly associated with plaque, with or without normalization using a normalization device.

[0207] In some embodiments, the system is configured to automatically and / or dynamically determine one or more vascular morphological parameters and / or plaque parameters from the medical images at block 208. In some embodiments, the one or more vascular morphological parameters and / or plaque parameters can include quantified parameters derived from the medical images. For example, in some embodiments, the system can be configured to utilize AI and / or ML algorithms or other algorithms to determine the one or more vascular morphological parameters and / or plaque parameters. As another example, in some embodiments, the system can be configured to determine one or more vascular morphological parameters, such as a classification of arterial remodeling due to plaque, which can further include positive arterial remodeling, negative arterial remodeling, and / or intermediate arterial remodeling. In some embodiments, the classification of arterial remodeling is determined based on a ratio of a maximum vessel diameter in the plaque region to a normal reference vessel diameter in the same region, which can be retrieved from a regular database. In some embodiments, the system can be configured to classify arterial remodeling as positive when a ratio of a maximum vessel diameter in the plaque region to a normal reference vessel diameter in the same region exceeds 1.1. In some embodiments, the system can be configured to classify arterial remodeling as negative when the ratio of the maximum vessel diameter in the plaque region to the normal reference vessel diameter is less than 0.95. In some embodiments, the system can be configured to classify arterial remodeling as intermediate when the ratio of the maximum vessel diameter in the plaque region to the normal reference vessel diameter is between 0.95 and 1.1.

[0208] Additionally, as part of block 208, in some embodiments, the system may be configured to determine the geometry and / or volume of one or more plaque regions and / or one or more blood vessels or arteries in block 201. For example, the system may be configured to determine whether the geometry of a particular plaque region is circular or elliptical or other shape. In some embodiments, the geometry of the plaque region may be a factor in assessing plaque stability. As another example, in some embodiments, the system may be configured to determine the curvature, diameter, length, volume, and / or any other parameter of a blood vessel or artery from a medical image.

[0209] In some embodiments, as part of block 208, the system may be configured to determine the volume and / or surface area of ​​the plaque region and / or the volume to surface area ratio or other function of the plaque region in block 203, such as, for example, the diameter, radius, and / or thickness of the plaque region. In some embodiments, a plaque having a small volume to surface area ratio may indicate that the plaque is stable. Thus, in some embodiments, the system may be configured to determine that a volume to surface area ratio of the plaque region that is less than a predetermined threshold is indicative of a stable plaque.

[0210] In some embodiments, as part of block 208, the system may be configured to determine a heterogeneity index of the plaque region in block 205. For example, in some embodiments, a plaque having low heterogeneity or high homogeneity may indicate that the plaque is stable. Thus, in some embodiments, the system may be configured to determine that heterogeneity of the plaque region that is below a predetermined threshold indicates a stable plaque. In some embodiments, the heterogeneity or homogeneity of the plaque region may be determined based on the heterogeneity or homogeneity of the radiodensity values ​​within the plaque region. Thus, in some embodiments, the system may be configured to determine the heterogeneity index of the plaque by generating a spatial mapping, such as a three-dimensional histogram, of the radiodensity values ​​within or across the geometry of the plaque region. In some embodiments, if the gradient or change in the radiodensity values ​​across the spatial mapping exceeds a certain threshold, the system may be configured to assign a high heterogeneity index. Conversely, in some embodiments, if the gradient or change in the radiodensity values ​​across the spatial mapping is below a certain threshold, the system may be configured to assign a low heterogeneity index.

[0211] In some embodiments, as part of block 208, the system may be configured to determine the radiodensity of the plaque and / or its composition in block 207. For example, a high radiodensity value may indicate that the plaque is highly calcified or stable, and a low radiodensity value may indicate that the plaque is less calcified or unstable. Thus, in some embodiments, the system may be configured to determine that a radiodensity of the plaque region that is above a predetermined threshold indicates a stabilized stable plaque. Additionally, different areas within the plaque region may be calcified to different levels, thereby exhibiting different radiodensity values. Thus, in some embodiments, the system may be configured to determine the radiodensity value of the plaque region, and / or the composition or percentage or change of radiodensity values ​​within the plaque region. For example, in some embodiments, the system may be configured to determine to what extent or what percentage of plaque within the plaque region exhibits radiodensity values ​​within the low range, medium range, high range, and / or any other classification.

[0212] Similarly, in some embodiments, as part of block 208, the system may be configured to determine a ratio of the radiodensity value of the plaque to the volume of the plaque in block 209. For example, it may be important to evaluate whether large or small areas of plaque exhibit high or low radiodensity values. Thus, in some embodiments, the system may be configured to determine the percentage composition of plaque that includes different radiodensity values ​​as a function or ratio of the volume of the plaque.

[0213] In some embodiments, as part of block 208, the system may be configured to determine the diffusivity and / or assign a diffusivity index to the plaque region in block 211. For example, in some embodiments, the diffusivity of the plaque may depend on the radiodensity value of the plaque, where a high radiodensity value may indicate low diffusivity or stability of the plaque.

[0214] In some embodiments, at block 210, the system may be configured to classify one or more plaque regions identified from the medical images as stable versus unstable or good versus bad based on one or more vascular morphological parameters and / or quantified plaque parameters determined and / or derived from the raw medical images. In particular, in some embodiments, the system may be configured to generate a weighted measure of one or more vascular morphological parameters and / or quantified plaque parameters determined and / or derived from the raw medical images. For example, in some embodiments, the system may be configured to weight one or more vascular morphological parameters and / or quantified plaque parameters equally. In some embodiments, the system may be configured to weight one or more vascular morphological parameters and / or quantified plaque parameters differently. In some embodiments, the system may be configured to weight one or more vascular morphological parameters and / or quantified plaque parameters logarithmically, algebraically, and / or utilizing another mathematical transformation. In some embodiments, the system may be configured to classify one or more plaque regions using the weighted criteria generated in block 210 and / or using only some of the vascular morphology parameters and / or quantified plaque parameters.

[0215] In some embodiments, at block 212, the system is configured to generate a quantized color mapping based on the analyzed and / or determined parameters. For example, in some embodiments, the system is configured to generate a visualization of the analyzed medical image by generating a quantized color mapping of calcified plaque, non-calcified plaque, benign plaque, malignant plaque, stable plaque, and / or unstable plaque, as determined using any of the analysis techniques described herein. Further, in some embodiments, the quantified color mapping can also include arterial and / or epicardial fat, which can also be determined by the system, for example, by utilizing one or more AI and / or ML algorithms.

[0216] In some embodiments, at block 214, the system is configured to generate a proposed treatment plan for the subject based on the analysis, e.g., a classification of plaque automatically derived from the raw medical images. In particular, in some embodiments, the system may be configured to assess or predict the subject's risk of atherosclerosis, stenosis, and / or ischemia based on the raw medical images and their automated image processing.

[0217] In some embodiments, one or more of the processes described herein in connection with Figure 2A may be repeated. For example, if medical images of the same subject are taken again at a later time, one or more of the processes described herein may be repeated and the analysis results may be used for disease tracking and / or other purposes.

[0218] Determination of noncalcified plaque from non-contrast CT images As discussed herein, in some embodiments, the system may be configured to utilize CT or other medical images of the subject as input to perform one or more image analysis techniques to assess the subject, including, for example, the risk of a cardiovascular event. In some embodiments, such CT images may include contrast-enhanced CT images, in which case some of the analysis techniques described herein may be directly applied, for example, to identify or classify plaque. However, in some embodiments, such CT images may include non-contrast CT images, in which case non-calcified plaque may be more difficult to identify and / or determine due to low radiodensity values ​​and overlap with other low radiodensity value components, such as blood. Thus, in some embodiments, the systems, devices, and methods described herein provide a novel approach to determining non-calcified plaque from non-contrast CT images that may be more widely available.

[0219] Also, in some embodiments, in addition to or instead of analyzing contrast-enhanced CT scans, the system can also be configured to examine attenuation densities in arteries that are lower than the attenuation density of blood flowing in the arteries in non-contrast CT scans. In some embodiments, these "low attenuation" plaques may be differentiated between blood attenuation density and fat, which may surround the coronary arteries and / or may represent non-calcified plaques of different materials. In some embodiments, the presence of these non-calcified plaques may provide an incremental prediction of whether already calcified plaques are stable or worsening or progressing or disappearing. These findings, measurable through these embodiments, may be linked to the patient's prognosis, with calcium stabilization (i.e., higher attenuation density) and absence of non-calcified plaques being associated with a favorable prognosis, and absence of calcium stabilization (i.e., no increase in attenuation density) or significant progression or new calcium formation being associated with a poor prognosis, including rapid progression of the disease, risk of heart attack, or other major adverse cardiovascular events.

[0220] FIG. 2B is a flow chart illustrating an overview of an example embodiment of a method for determining non-calcified and / or low-attenuating plaque from a medical image, such as a non-contrast CT image. As discussed herein and illustrated in FIG. 2B, in some embodiments, a system can be configured to determine non-calcified and / or low-attenuating plaque from a medical image. In some embodiments, the medical image can be of a coronary artery region of a subject or patient. In some embodiments, the medical image can be obtained using one or more modalities, such as CT, dual energy computed tomography (DECT), spectral CT, X-ray, ultrasound, echocardiography, IVUS, MR, OCT, nuclear medicine imaging, PET, SPECT, NIRS, and / or others. In some embodiments, the system can be configured to access one or more medical images, for example from the medical image database 100, at block 202.

[0221] In some embodiments, to determine non-calcified and / or low-attenuating plaque from a medical image or non-contrast CT image, the system may be configured to utilize a staged approach to first identify areas within the medical image that are clearly non-calcified plaque. In some embodiments, the system may then perform a more detailed analysis of the remaining areas in the image to identify other regions of non-calcified and / or low-attenuating plaque. By utilizing such a compartmentalized or staged approach, in some embodiments, the system may identify or determine non-calcified and / or low-attenuating plaque from a medical image or non-contrast CT image with a quicker turnaround than if a more complex analysis had to be applied to every region or pixel of the image.

[0222] In particular, in some embodiments, at block 224, the system may be configured to identify epicardial fat from the medical image. In some embodiments, the system may be configured to identify epicardial fat by determining all pixels or regions in the image having radiodensity values ​​below a predefined threshold and / or within a predefined range. The exact predefined threshold value or range of radiodensity for identifying epicardial fat may depend on the medical image, the type of scanner, the scan parameters, and / or otherwise, and for that reason, in some examples, a normalization device may be used to normalize the medical image. For example, in some embodiments, the system may be configured to identify pixels and / or regions of epicardial fat in the medical image or non-contrast CT image having radiodensity values ​​within a range that includes about -100 Hounsfield units and / or -100 Hounsfield units. In particular, in some embodiments, the system provides a lower limit of about -100 Hounsfield units, about -110 Hounsfield units, about -120 Hounsfield units, about -130 Hounsfield units, about -140 Hounsfield units, about -150 Hounsfield units, about -160 Hounsfield units, about -170 Hounsfield units, about -180 Hounsfield units, about -190 Hounsfield units, or about -200 Hounsfield units, and a lower limit of about 30 Hounsfield units, about 20 Hounsfield units, about 10 Hounsfield units, or a lower limit of about 20 Hounsfield units, about 30 Hounsfield units, about 40 Hounsfield units, about 50 Hounsfield units, about 60 Hounsfield units, about 70 Hounsfield units, about 80 Hounsfield units, about 90 Hounsfield units, or about 100 Hounsfield units. The method may be configured to identify epicardial fat pixels and / or regions within the medical image or non-contrast CT image as having radio density values ​​within a range having an upper limit of about 0 Hounsfield units, about 0 Hounsfield units, about -10 Hounsfield units, about -20 Hounsfield units, about -30 Hounsfield units, about -40 Hounsfield units, about -50 Hounsfield units, about -60 Hounsfield units, about -70 Hounsfield units, about -80 Hounsfield units, or about -90 Hounsfield units.

[0223] In some embodiments, the system can be configured to identify and / or segment arteries on medical images or non-contrast CT images using the identified epicardial fat as the outer boundary of the arteries. For example, the system can be configured to first identify regions of epicardial fat on the medical images and assign the volume between the epicardial fat as an artery, such as a coronary artery.

[0224] In some embodiments, at block 226, the system may be configured to identify a first set of pixels or regions in the medical image, such as within the identified artery, as non-calcified or low-attenuating plaque. More specifically, in some embodiments, the system may be configured to identify the initial set of low-attenuating or non-calcified plaque by identifying pixels or regions having radiodensity values ​​below a predetermined threshold or within a predetermined range. For example, the predetermined threshold or predetermined range may be set to allow the resulting pixels to be reliably marked as low-attenuating or non-calcified plaque that is not likely to be confused with another material, such as blood. In particular, in some embodiments, the system may be configured to identify the initial set of low-attenuating or non-calcified plaque by identifying pixels or regions having radiodensity values ​​below about 30 Hounsfield units. In some embodiments, the system may be configured to identify a first set of low-attenuating or non-calcified plaques by identifying pixels or regions having radiodensity values ​​of about 60 Hounsfield units, about 55 Hounsfield units, about 50 Hounsfield units, about 45 Hounsfield units, about 40 Hounsfield units, about 35 Hounsfield units, about 30 Hounsfield units, about 25 Hounsfield units, about 20 Hounsfield units, about 15 Hounsfield units, about 10 Hounsfield units, about 5 Hounsfield units or less and / or having radiodensity values ​​of about 0 Hounsfield units, about 5 Hounsfield units, about 10 Hounsfield units, about 15 Hounsfield units, about 20 Hounsfield units, about 25 Hounsfield units, and / or about 30 Hounsfield units or more. In some embodiments, the system may be configured to classify pixels or regions that are within or below this predetermined range of radiodensity values ​​as the first set of identified non-calcified or low-attenuating plaques at block 238.

[0225] In some embodiments, the system may be configured to identify a second set of pixels or regions in the medical image, such as within the identified artery, that may or may not represent low attenuating or non-calcified plaque, at block 228. As noted above, in some embodiments, this second set of candidate pixels or regions may require additional analysis to confirm that they represent plaque. In particular, in some embodiments, the system may be configured to identify a second set of pixels or regions that may be low attenuating or non-calcified plaque by identifying pixels or regions of the image that have radiodensity values ​​within a predetermined range. In some embodiments, the predetermined range for identifying this second set of pixels or regions may be between about 30 Hounsfield units and 100 Hounsfield units. In some embodiments, the predetermined range for identifying this second set of pixels or regions may have a lower limit of about 0 Hounsfield units, 5 Hounsfield units, 10 Hounsfield units, 15 Hounsfield units, 20 Hounsfield units, 25 Hounsfield units, 30 Hounsfield units, 35 Hounsfield units, 40 Hounsfield units, 45 Hounsfield units, 50 Hounsfield units, and / or an upper limit of about 55 Hounsfield units, 60 Hounsfield units, 65 Hounsfield units, 70 Hounsfield units, 75 Hounsfield units, 80 Hounsfield units, 85 Hounsfield units, 90 Hounsfield units, 95 Hounsfield units, 100 Hounsfield units, 110 Hounsfield units, 120 Hounsfield units, 130 Hounsfield units, 140 Hounsfield units, 150 Hounsfield units.

[0226] In some embodiments, at block 230, the system may be configured to perform an analysis of the heterogeneity of the identified second set of pixels or regions. For example, depending on the range of radiodensity values ​​used to identify the second set of pixels, in some embodiments, the second set of pixels or regions may include blood and / or plaque. Blood may generally exhibit a more homogeneous gradient of radiodensity values ​​compared to plaque. Thus, in some embodiments, by analyzing the homogeneity or heterogeneity of the pixels or regions identified as part of the second set, the system may be able to distinguish between blood and non-calcified or low-attenuating plaque. Thus, in some embodiments, the system may be configured to determine a heterogeneity index in the second set of regions of pixels identified from the medical image by generating a spatial mapping, such as a three-dimensional histogram, of radiodensity values ​​within or across the geometry or region of the plaque. In some embodiments, if the gradient or change in radiodensity values ​​across the spatial mapping exceeds a certain threshold, the system may be configured to assign a high heterogeneity index and / or classify as plaque. Conversely, in some embodiments, if the gradient or change in radiodensity values ​​across the spatial mapping is below a certain threshold, the system can be configured to assign a low heterogeneity index and / or classify as blood.

[0227] In some embodiments, at block 240, the system may be configured to identify a subset of the second set of regions of identified pixels from the medical image as plaque or non-calcified or low-attenuating plaque. In some embodiments, at block 242, the system may be configured to combine the first set of identified non-calcified or low-attenuating plaque from block 238 with the second set of identified non-calcified or low-attenuating plaque from block 240. Thus, even with the use of non-contrast CT images, in some embodiments the system may be configured to identify low-attenuating or non-calcified plaque, which may be more difficult to identify compared to calcified or high-attenuating plaque due to possible overlap with other material such as blood.

[0228] In some embodiments, the system may also be configured to determine calcified or hyper-attenuating plaque from the medical image at block 232. This process may be simpler compared to identifying hypo-attenuating or non-calcified plaque from the medical image or non-contrast CT image. In particular, in some embodiments, the system may be configured to identify calcified or hyper-attenuating plaque from the medical image or non-contrast CT image by identifying pixels or regions in the image that have radiodensity values ​​above a predetermined threshold and / or within a predetermined range. For example, in some embodiments, the system may provide a signal of about 100 Hounsfield units, about 150 Hounsfield units, about 200 Hounsfield units, about 250 Hounsfield units, about 300 Hounsfield units, about 350 Hounsfield units, about 400 Hounsfield units, about 450 Hounsfield units, about 500 Hounsfield units, about 600 Hounsfield units, about 700 Hounsfield units, about 800 Hounsfield units, about 900 Hounsfield units, about 1000 Hounsfield units, about 1100 Hounsfield units, about 1200 Hounsfield units, about 1300 Hounsfield units, about 1400 Hounsfield units, about 1500 Hounsfield units, about 1600 Hounsfield units, about 1700 Hounsfield units, about 1800 Hounsfield units, about 1900 Hounsfield units, about 2100 Hounsfield units, about 2200 Hounsfield units, about 2300 Hounsfield units, about 2400 Hounsfield units, about 2500 Hounsfield units, about 300 Hounsfield units, about 350 Hounsfield units, about 400 Hounsfield units, about 450 Hounsfield units, about 500 Hounsfield units, about 600 Hounsfield units, about 700 Hounsfield units, about 800 Hounsfield units, about 900 Hounsfield units, about 1000 Hounsfield units, about 1100 Hounsfield units, about 1200 Hounsfield units, about 2500 Hounsfield units, about 3500 Hounsfield units, about 3600 Hounsfield units, about 3 In one embodiment, the present invention can be configured to identify as calcification or high attenuation plaque regions or pixels from medical images or non-contrast CT images having radiodensity values ​​above about 100 Hounsfield units, about 1300 Hounsfield units, about 1400 Hounsfield units, about 1500 Hounsfield units, about 1600 Hounsfield units, about 1700 Hounsfield units, about 1800 Hounsfield units, about 1900 Hounsfield units, about 2000 Hounsfield units, about 2500 Hounsfield units, about 3000 Hounsfield units, and / or any other minimum threshold.

[0229] In some embodiments, at block 234, the system may be configured to generate a quantized color mapping of one or more identified materials from the medical image. For example, in some embodiments, the system may be configured to assign different colors to different regions associated with different materials, such as non-calcified or low-attenuating plaque, calcified or high-attenuating plaque, all plaque, arteries, epicardial fat, and / or others. In some embodiments, the system may be configured to generate a visualization of the quantized color map and / or present it to a medical professional or patient via a GUI. In some embodiments, at block 236, the system may be configured to generate a suggested treatment plan for the disease based on one or more of the identified non-calcified or low-attenuating plaque, calcified or high-attenuating plaque, all plaque, arteries, epicardial fat, and / or others. For example, in some embodiments, the system may be configured to generate treatment plans for arterial disease, renal artery disease, abdominal atherosclerosis, carotid atherosclerosis, and / or the like, and the medical images analyzed may be taken from any one or more areas of the subject for analysis of such diseases.

[0230] In some embodiments, one or more of the processes described herein in connection with Figure 2B can be repeated. For example, if medical images of the same subject are taken again at a later time, one or more of the processes described herein can be repeated and the analysis results can be used for disease tracking and / or other purposes.

[0231] Additionally, in some embodiments, the system may be configured to identify and / or determine non-calcified plaque from a DECT or spectral CT image. Similar to the process described above, in some embodiments, the system may be configured to access a DECT or spectral CT image, identify epicardial fat on the DECT image or spectral CT, and / or segment one or more arteries on the DECT image or spectral CT, and identify and / or classify a first set of pixels or regions within the artery as a first set of low-attenuation or non-calcified plaque, and / or identify a second set of pixels or regions within the artery as a second set of low-attenuation or non-calcified plaque. However, unlike the techniques described above, in some embodiments, such as when a DECT or spectral CT image is analyzed, the system may be configured to identify a subset of the second set of pixels without having to perform a heterogeneity and / or homogeneity analysis of the second set of pixels. Instead, in some embodiments, the system may be configured to distinguish between blood and low-attenuation or non-calcified plaque directly from the image, for example, by utilizing dual or multi-spectral aspects of the DECT or spectral CT image. In some embodiments, the system may be configured to combine a subset of the first set of identified pixels or regions with a second set of pixels or regions identified as low attenuation or non-calcified plaque and identify the entire set on the medical image. In some embodiments, even when analyzing DECT or spectral CT images, the system may be configured to further analyze the second set of pixels or regions by performing a heterogeneity or homogeneity analysis, similar to that described above in connection with block 230. For example, even when analyzing DECT or spectral CT images, in some embodiments, the distinction between certain areas of blood and / or low attenuation or non-calcified plaque may not be complete and / or accurate.

[0232] Imaging-based risk assessment In some embodiments, the systems, devices, and methods described herein are configured to utilize medical image-based processing to assess a subject's risk for cardiovascular events, major adverse cardiovascular events (MACE), rapid plaque progression, and / or non-response to drugs. In particular, in some embodiments, the systems can be configured to automatically and / or dynamically assess such health risks of a subject by analyzing only non-invasively obtained medical images, e.g., using AI and / or ML algorithms, and provide a full image-based analysis report within minutes.

[0233] In particular, in some embodiments, the system can be configured to calculate the total amount of plaque (and / or the amount of a particular type of plaque) in a particular artery and / or in all arteries of the patient. In some embodiments, the system can be configured to determine the total amount of malignant plaque in a particular artery and / or in the entire arterial area across some or all of the arteries of the patient. In some embodiments, the system can be configured to determine a risk factor and / or diagnosis for a particular patient to suffer a heart attack or other cardiac event based on the total amount of plaque in a particular artery and / or in the entire arterial area across some or all of the arteries of the patient. Other risk factors that can be determined from the amount of "malignant" plaque, or the relative amount of "malignant" plaque to "benign" plaque, can include the rate of disease progression and / or the likelihood of ischemia. In some embodiments, plaque can be measured by total volume (or area in cross-sectional imaging), as well as by relative amount when normalized to total vessel volume, total vessel length, or subtended myocardium.

[0234] In some embodiments, the coronary imaging data may include measures of atherosclerosis, stenosis, and vascular morphology. In some embodiments, this information may be combined with other cardiovascular disease phenotype extractions by quantitative characterization of the left and right ventricles, the left and right atria, the aortic, mitral, tricuspid, and pulmonary valves, the aorta, pulmonary arteries, pulmonary veins, coronary sinuses, and the inferior and superior vena cava, epicardial or pericoronary fat, lung density, bone density, pericardium, and other. As an example, in some embodiments, the coronary imaging data may be integrated with the left ventricular mass and segmented according to the amount and location of the artery it faces. This combination of the left ventricular partial myocardial mass and coronary information may improve prediction of whether future heart attacks will be large or small. As another example, in some embodiments, the coronary vascular volume may be related to the left ventricular mass as a measure of left ventricular hypertrophy, which may be commonly found in hypertensive patients. Increased left ventricular mass (relative or absolute) may indicate worsening disease or uncontrolled hypertension. As another example, in some embodiments, the onset, progression, and / or worsening of atrial fibrillation may be predicted by atrial size, volume, atrial free wall mass and thickness, atrial function, and fat surrounding the atria. In some embodiments, these predictions may be made using ML or AI algorithms or other algorithm types.

[0235] In succession, in some embodiments, algorithms enabling segmentation of atherosclerosis, stenosis, and vascular morphology, along with those enabling segmentation of other cardiovascular and thoracic structures, may serve as inputs to prognostic algorithms. In some embodiments, the output of the prognostic algorithms, or those enabling image segmentation, may be leveraged as inputs to other algorithms, which may then guide clinical decision-making by predicting future events. As an example, in some embodiments, the combined scoring of atherosclerosis, stenosis, and / or vascular morphology may identify patients who may benefit from coronary revascularization, i.e., patients who will achieve symptom relief and have a reduced risk of heart attack and death. As another example, in some embodiments, integrated scoring of atherosclerosis, stenosis, and vascular morphology may identify individuals who may benefit from certain types of medications, such as lipid-lowering drugs (such as statins, PCSK-9 inhibitors, ethyl icosapentate, and others), Lp(a)-lowering drugs, and antithrombotic drugs (such as clopidogrel, rivaroxaban, and others). In some embodiments, the benefit predicted by these algorithms may be a determination of the type of plaque progression (progression, resolution, or mixed response), stabilization with medication, and / or the need for enhanced aggressive therapy to reduce progression. In some embodiments, imaging data may be combined with other data to identify areas within the coronary vessels that are normal and currently free of plaque, but where future plaque formation is likely.

[0236] In some embodiments, automated or manual co-registration methods can be combined with imaging segmentation data to compare two or more images over time. In some embodiments, comparison of these images can allow for the determination of differences in coronary atherosclerosis, stenosis, and vascular morphology over time, which can be used as input variables for risk prediction.

[0237] In some embodiments, coronary imaging data regarding atherosclerosis, stenosis, and vessel morphology, with or without being combined with thoracic and cardiovascular disease measurements, can be integrated into algorithms that determine whether coronary vessels are ischemic or exhibiting reduced blood flow or blood pressure (in either resting or hyperemic states).

[0238] In some embodiments, the algorithms for coronary atherosclerosis, stenosis, and ischemia can be modified by the computer system and / or otherwise to remove or "seal" plaque. In some embodiments, a comparison can be made before and after the system removes or seals plaque to determine whether any changes have occurred. For example, in some embodiments, the system can be configured to determine whether coronary ischemia has been removed by sealing the plaque.

[0239] In some embodiments, characterization of coronary atherosclerosis, stenosis, and / or vascular morphology can allow for correlation of a patient's biological age with vascular age when compared to a population-based cohort of similarly scanned patients. As an example, a 60 year old patient may have X units of plaque in their coronary arteries equivalent to an average 70 year old patient in a population-based cohort. In this case, the patient's vascular age may be 10 years older than the patient's biological age.

[0240] In some embodiments, risk assessment enabled by image segmentation predictive algorithms may allow refinement of measures of disease or mortality probability in people considered for disability or life insurance. In this scenario, risk assessment may replace or augment traditional actuarial algorithms.

[0241] In some embodiments, the imaging data may be combined with other data to enhance risk assessment for future adverse events such as heart attack, stroke, death, rapid progression, non-response to drug therapy, no reflow phenomenon, etc. In some embodiments, the other data may include multi-omics approaches, where algorithms integrate imaging phenotype data with genotype data, proteomic data, transcriptomic data, metabolomic data, microbiomics data, and / or activity and lifestyle data, such as measured by a smartphone or similar device.

[0242] FIG. 3A is a flow chart illustrating an overview of an example embodiment of a method for risk assessment based on medical image analysis. As shown in FIG. 3A, in some embodiments, the system can be configured to access medical images at block 202. Further, in some embodiments, the system can be configured to identify one or more arteries at block 204 and / or one or more plaque regions at block 206. Additionally, in some embodiments, the system can be configured to determine one or more vessel morphology and / or quantified plaque parameters at block 208 and / or classify stable or unstable plaque based on the determined one or more vessel morphology and / or quantified plaque parameters and / or weighted criteria thereof at block 210. Additional details regarding the processes and techniques represented by blocks 202, 204, 206, 208, and 210 can be found in the above description regarding FIG. 2A.

[0243] In some embodiments, the system can, for example, use the classified stable and / or unstable plaque regions to automatically and / or dynamically determine and / or generate a risk of a cardiovascular event for the subject in block 302. More specifically, in some embodiments, the system can utilize AI, ML, or other algorithms to generate a risk of a cardiovascular event, MACE, rapid plaque progression, and / or non-response to a drug in block 302 based on the image analysis.

[0244] In some embodiments, at block 304, the system can be configured to compare the determined one or more vascular morphology parameters, quantified plaque parameters, and / or classified stable vs. unstable plaque, and / or values ​​thereof, such as volumes, ratios, and / or the like, to one or more known datasets of coronary artery values ​​derived from one or more other subjects. The one or more known datasets can include one or more vascular morphology parameters, quantified plaque parameters, and / or classified stable vs. unstable plaque, and / or values ​​thereof, such as volumes, ratios, and / or the like, derived from medical images taken from other subjects, including healthy subjects and / or subjects with various risk levels. For example, the one or more known datasets of coronary artery values ​​can be stored in a coronary artery values ​​database 306, which can be locally accessible by the system and / or remotely accessible by the system via a network connection.

[0245] In some embodiments, at block 308, the system can be configured to update a risk of a cardiovascular event for the subject based on the comparison against one or more known datasets. For example, based on the comparison, the system may increase or decrease a previously generated risk assessment. In some embodiments, the system may maintain the previously generated risk assessment even after the comparison. In some embodiments, the system can be configured to generate a suggested treatment for the subject based on the generated and / or updated risk assessment after the comparison against the known dataset of coronary artery values.

[0246] In some embodiments, the system may be configured to further identify and / or determine one or more parameters associated with one or more other cardiovascular structures from the medical image at block 310. For example, the one or more additional cardiovascular structures may include the left ventricle, the right ventricle, the left atrium, the right atrium, the aortic valve, the mitral valve, the tricuspid valve, the pulmonary valve, the aorta, the pulmonary artery, the inferior and superior vena cava, epicardial fat, and / or the pericardium.

[0247] In some embodiments, parameters associated with the left ventricle may include size, mass, volume, shape, eccentricity, surface area, thickness, and / or the like. Similarly, in some embodiments, parameters associated with the right ventricle may include size, mass, volume, shape, eccentricity, surface area, thickness, and / or the like. In some embodiments, parameters associated with the left atrium may include size, mass, volume, shape, eccentricity, surface area, thickness, pulmonary vein angulation, atrial appendage morphology, and / or the like. In some embodiments, parameters associated with the right atrium may include size, mass, volume, shape, eccentricity, surface area, thickness, and / or the like.

[0248] Additionally, in some embodiments, parameters associated with the aortic valve may include thickness, volume, mass, calcification, a three-dimensional map of calcification and density, eccentricity of calcification, classification by individual leaflets, and / or others. In some embodiments, parameters associated with the mitral valve may include thickness, volume, mass, calcification, a three-dimensional map of calcification and density, eccentricity of calcification, classification by individual leaflets, and / or others. In some embodiments, parameters associated with the tricuspid valve may include thickness, volume, mass, calcification, a three-dimensional map of calcification and density, eccentricity of calcification, classification by individual leaflets, and / or others. Additionally, in some embodiments, parameters associated with the pulmonary valve may include thickness, volume, mass, calcification, a three-dimensional map of calcification and density, eccentricity of calcification, classification by individual leaflets, and / or others.

[0249] In some embodiments, parameters associated with the aorta may include dimensions, volume, diameter, area, enlargement, protrusion, and / or the like. In some embodiments, parameters associated with the pulmonary artery may include dimensions, volume, diameter, area, enlargement, protrusion, and / or the like. In some embodiments, parameters associated with the inferior vena cava and superior vena cava may include dimensions, volume, diameter, area, enlargement, protrusion, and / or the like.

[0250] In some embodiments, parameters associated with epicardial fat can include volume, density, three-dimensional density, and / or the like, in some embodiments, parameters associated with the pericardium can include thickness, mass, and / or the like.

[0251] In some embodiments, the system may be configured to classify one or more of the other identified cardiovascular structures, e.g., using the one or more determined parameters, at block 312. In some embodiments, with respect to one or more of the other identified cardiovascular structures, the system may be configured to classify each as normal versus abnormal, increasing or decreasing, and / or static or dynamic over time.

[0252] In some embodiments, at block 314, the system can be configured to compare the determined one or more parameters of the other cardiovascular structures with one or more known datasets of cardiovascular structural parameters derived from one or more other subjects. The one or more known datasets of cardiovascular structural parameters can include any one or more of the parameters described above associated with the other cardiovascular structures. In some embodiments, the one or more known datasets of cardiovascular structural parameters can be derived from medical images taken from other subjects, including healthy subjects and / or subjects at various risk levels. In some embodiments, the one or more known datasets of cardiovascular structural parameters can be stored in a cardiovascular structural value or cardiovascular disease (CVD) database 316, which can be locally accessible by the system and / or remotely accessible by the system via a network connection.

[0253] In some embodiments, at block 318, the system may be configured to update the risk of a cardiovascular event for the subject based on a comparison of the cardiovascular structural parameters to one or more known data sets. For example, based on the comparison, the system may increase or decrease a previously generated risk assessment. In some embodiments, the system may maintain the previously generated risk assessment even after the comparison.

[0254] In some embodiments, at block 320, the system may be configured to generate a quantified color map, which may include color coding for one or more other cardiovascular structures identified from the medical images, stable plaque, unstable plaque, arteries, and / or other. In some embodiments, at block 322, the system may be configured to generate a suggested treatment for the subject based on the generated and / or updated risk assessment after comparison of the cardiovascular structural parameters to known datasets.

[0255] In some embodiments, at block 324, the system may be configured to further identify one or more non-cardiovascular structures from the medical image and / or determine one or more parameters associated therewith. For example, the medical image may include one or more non-cardiovascular structures within the field of view. In particular, the one or more non-cardiovascular structures may include lungs, bones, liver, and / or others.

[0256] In some embodiments, parameters associated with non-cardiovascular structures may include volume, surface area, volume to surface area ratio or function, non-uniformity of radiodensity values, radiodensity values, geometric shape (elliptical, spherical, and / or other), spatial radiodensity, spatial scarring, and / or other. Additionally, in some embodiments, parameters associated with lungs may include density, scarring, and / or other. For example, in some embodiments, the system may be configured to associate low Hounsfield units in an area of ​​the lung with emphysema. In some embodiments, parameters associated with bones, such as the spine and / or ribs, may include radiodensity, presence and / or degree of fracture, and / or other. For example, in some embodiments, the system may be configured to associate low Hounsfield units in an area of ​​bone with osteoporosis. In some embodiments, parameters associated with the liver may include density of non-alcoholic fatty liver disease, which may be assessed by the system by analyzing and / or comparing with the Hounsfield unit density of the liver.

[0257] In some embodiments, the system may be configured to classify one or more of the identified non-cardiovascular structures, e.g., using the one or more determined parameters, at block 326. In some embodiments, with respect to one or more of the identified non-cardiovascular structures, the system may be configured to classify each as normal versus abnormal, increasing or decreasing, and / or static or dynamic over time.

[0258] In some embodiments, at block 328, the system can be configured to compare the determined one or more parameters of the non-cardiovascular structures with one or more known data sets of non-cardiovascular structural parameters or non-CVD values ​​derived from one or more other subjects. The one or more known data sets of non-cardiovascular structural parameters or non-CVD values ​​can include any one or more of the parameters described above associated with the non-cardiovascular structures. In some embodiments, the one or more known data sets of non-cardiovascular structural parameters or non-CVD values ​​can be derived from medical images taken from other subjects, including healthy subjects and / or subjects at various risk levels. In some embodiments, the one or more known data sets of non-cardiovascular structural parameters or non-CVD values ​​can be stored in a non-cardiovascular structural values ​​or non-CVD database 330, which can be locally accessible by the system and / or remotely accessible by the system via a network connection.

[0259] In some embodiments, at block 332, the system may be configured to update the risk of a cardiovascular event for the subject based on a comparison of the non-cardiovascular structural parameters or non-CVD values ​​to one or more known data sets. For example, based on the comparison, the system may increase or decrease a previously generated risk assessment. In some embodiments, the system may maintain the previously generated risk assessment even after the comparison.

[0260] In some embodiments, at block 334, the system can be configured to generate a quantified color map, which can include color coding for one or more non-cardiovascular structures identified from the medical image, as well as other cardiovascular structures identified from the medical image, stable plaque, unstable plaque, arteries, and / or others. In some embodiments, at block 336, the system can be configured to generate a suggested treatment for the subject based on the generated and / or updated risk assessment after comparison with known datasets of non-cardiovascular structural parameters or non-CVD values.

[0261] In some embodiments, one or more of the processes described herein in connection with Figure 3A may be repeated. For example, if medical images of the same subject are taken again at a later time, one or more of the processes described herein may be repeated and the analysis results may be used to track the subject's risk assessment based on the image processing, and / or for other purposes.

[0262] Quantification of atherosclerosis In some embodiments, the system is configured to automatically and / or dynamically quantify atherosclerosis by analyzing one or more arteries present in a medical image, such as CT scan data. In some embodiments, the system is configured to quantify atherosclerosis as a course of primary disease, while stenosis and / or ischemia can be considered as surrogates. Prior to the embodiments described herein, quantification of primary disease was not feasible due to the lengthy manual process and labor required to perform the process, which can take 4-8 hours or more. In contrast, in some embodiments, the system is configured to quantify atherosclerosis based on analysis of medical images and / or CT scans using one or more AI, ML, and / or other algorithms that can segment, identify, and / or quantify atherosclerosis in less than about 1 minute, about 2 minutes, about 3 minutes, about 4 minutes, about 5 minutes, about 6 minutes, about 7 minutes, about 8 minutes, about 9 minutes, about 10 minutes, about 11 minutes, about 12 minutes, about 13 minutes, about 14 minutes, about 15 minutes, about 20 minutes, about 25 minutes, about 30 minutes, about 40 minutes, about 50 minutes, and / or about 60 minutes. In some embodiments, the system is configured to quantify atherosclerosis within a time frame defined by two of the above values. In some embodiments, the system is configured to calculate stenosis rather than simply visualizing it, thereby allowing the user to better understand the total cardiac atherosclerosis and / or ensuring that the calculated results of stenosis are the same when the same medical image is used for analysis. Importantly, the type of atherosclerosis can also be quantified and / or classified by this method. The type of atherosclerosis can be determined binary (calcified vs. non-calcified plaque), sequential (densely calcified plaque, calcified plaque, fibrous plaque, fibrofatty plaque, necrotic core, or a mixture of plaque types), or continuously (such as by attenuation density on the Hounsfield unit scale).

[0263] 3B is a flow chart illustrating an overview of an example embodiment of a method for quantifying and / or classifying atherosclerosis based on medical image analysis. As shown in FIG. 3B, in some embodiments, the system may be configured to access medical images, such as a CT scan of a subject's coronary artery region, at block 202. Further, in some embodiments, the system may be configured to identify one or more arteries at block 204 and / or one or more plaque regions at block 206. Additionally, in some embodiments, the system may be configured to determine one or more vascular morphology and / or quantified plaque parameters at block 208. For example, in some embodiments, the system may be configured to determine the geometry and / or volume of the region of plaque and / or blood vessel at block 201, a volume to surface area ratio or function of the region of plaque at block 203, a heterogeneity or homogeneity index of the region of plaque at block 205, the radiodensity of the region of plaque and / or its composition by a range of radiodensity values ​​at block 207, a radiodensity to volume ratio of the region of plaque at block 209, and / or a diffusivity of the region of plaque at block 211. Additional details regarding the processes and techniques represented by blocks 202, 204, 206, 208, 201, 203, 205, 207, 209, and 211 may be found in the discussion above with respect to FIG.

[0264] In some embodiments, the system may be configured to quantify and / or classify atherosclerosis based on the determined one or more vascular morphology and / or quantified plaque parameters, block 340. In some embodiments, the system may be configured to generate a weighted measure of the one or more vascular morphology parameters and / or quantified plaque parameters determined and / or derived from the raw medical images. For example, in some embodiments, the system may be configured to weight the one or more vascular morphology parameters and / or quantified plaque parameters equally. In some embodiments, the system may be configured to weight the one or more vascular morphology parameters and / or quantified plaque parameters differently. In some embodiments, the system may be configured to weight the one or more vascular morphology parameters and / or quantified plaque parameters logarithmically, algebraically, and / or utilizing another mathematical transformation. In some embodiments, the system may be configured to quantify and / or classify atherosclerosis using the weighted measure, block 340, and / or using only some of the vascular morphology parameters and / or quantified plaque parameters.

[0265] In some embodiments, the system is configured to generate a weighted measure of the one or more vascular morphological parameters and / or quantified plaque parameters by comparing with one or more known vascular morphological parameters and / or quantified plaque parameters derived from medical images of other subjects. For example, one or more known vascular morphological parameters and / or quantified plaque parameters can be derived from one or more healthy subjects and / or subjects at risk for coronary vascular disease.

[0266] In some embodiments, the system is configured to classify the subject's atherosclerosis as one or more of high risk, medium risk, or low risk based on the quantified atherosclerosis. In some embodiments, the system is configured to classify the subject's atherosclerosis based on the quantified atherosclerosis using AI, ML, and / or other algorithms. In some embodiments, the system is configured to classify the subject's atherosclerosis by combining and / or weighting one or more of volume-to-surface area ratio, volume, heterogeneity index, and radiodensity of one or more regions of plaque.

[0267] In some embodiments, a plaque with a small volume to surface area ratio or a small absolute volume itself can indicate that the plaque is stable. Thus, in some embodiments, the system can be configured to determine that a volume to surface area ratio of a plaque region that is below a predetermined threshold indicates a low risk of atherosclerosis. Thus, in some embodiments, the system can be configured to take into account the number and / or surface area of ​​the plaque. For example, if there are more plaques with smaller surfaces, it can be associated with a larger surface area or greater irregularity, which in turn can be associated with a larger surface area to volume ratio. In contrast, if there are fewer plaques with larger surfaces or greater regularity, it can be associated with a smaller surface area to volume ratio or a larger volume to surface area ratio. In some embodiments, a high radiodensity value can indicate that the plaque is highly calcified or stable, and a low radiodensity value can indicate that the plaque is less calcified or unstable. Thus, in some embodiments, the system can be configured to determine that a radiodensity of a plaque region that is above a predetermined threshold indicates a low risk of atherosclerosis. In some embodiments, a plaque with low heterogeneity or high homogeneity can indicate that the plaque is stable, and thus, in some embodiments, the system can be configured to determine that heterogeneity of a plaque region that is below a predetermined threshold indicates a low risk of atherosclerosis.

[0268] In some embodiments, the system is configured to calculate or determine a numerical calculation or representation of coronary artery stenosis based on the quantified and / or classified atherosclerosis derived from the medical images at block 342. In some embodiments, the system is configured to calculate the stenosis using one or more vascular morphology parameters and / or quantified plaque parameters derived from medical images of the subject's coronary artery regions.

[0269] In some embodiments, the system is configured to predict a risk of ischemia for the subject based on the quantified and / or classified atherosclerosis derived from the medical images, at block 344. In some embodiments, the system is configured to calculate the risk of ischemia using one or more vascular morphology parameters and / or quantified plaque parameters derived from medical images of the subject's coronary artery territories.

[0270] In some embodiments, the system is configured to generate a suggested treatment for the subject based on the quantified and / or classified atherosclerosis, stenosis, and / or ischemic risk, both derived automatically and / or dynamically from the raw medical images using image processing algorithms and techniques.

[0271] In some embodiments, one or more of the processes described herein in connection with Figure 3A may be repeated. For example, if medical images of the same subject are taken again at a later time, one or more of the processes described herein may be repeated and the analysis results may be used to track the subject's quantified atherosclerosis, and / or for other purposes.

[0272] Quantification of plaque, stenosis, and / or CAD-RADS score As discussed herein, in some embodiments, the system is configured to make inferences from the interpretation of medical images and provide substantially accurate and / or substantially precise calculations or estimates of the percentage of stenosis, atherosclerosis, and / or Coronary Artery Disease-Reporting and Data System (CAD-RADS) score as derived from the medical images. Thus, in some embodiments, the system can improve imager readings by providing comprehensive quantitative analysis that can improve efficiency, accuracy, and / or reproducibility.

[0273] 3C is a flow chart showing an overview of an example embodiment of a method for quantifying stenosis and generating a CAD-RADS score based on medical image analysis. As shown in FIG. 3A, in some embodiments, the system can be configured to access medical images at block 202. Additional details regarding the types of medical images and other processes and techniques represented in block 202 can be found in the description above regarding FIG. 2A.

[0274] In some embodiments, at block 354, the system is configured to identify one or more arteries, plaque, and / or fat in the medical image, e.g., using AI, ML, and / or other algorithms. The process and techniques for identifying one or more arteries, plaque, and / or fat may include one or more of the same features as described above with respect to blocks 204 and 206. In particular, in some embodiments, the system may be configured to automatically and / or dynamically identify one or more arteries, including, e.g., coronary arteries, carotid arteries, aorta, renal arteries, lower limb arteries, and / or cerebral arteries, utilizing one or more AI and / or ML algorithms. In some embodiments, the one or more AI and / or ML algorithms may be trained using a convolutional neural network (CNN) on a set of medical images in which arteries have been identified, thereby enabling the AI ​​and / or ML algorithm to automatically identify arteries directly from the medical images. In some embodiments, the arteries are identified by size and / or location.

[0275] Further, in some embodiments, the system can be configured to identify one or more plaque regions in a medical image, for example, using one or more AI and / or ML algorithms to automatically and / or dynamically identify one or more plaque regions. In some embodiments, the one or more AI and / or ML algorithms can be trained using a convolutional neural network (CNN) on a set of medical images in which plaque regions have been identified, thereby enabling the AI ​​and / or ML algorithm to automatically identify plaque regions directly from the medical images. In some embodiments, the system can be configured to identify a vessel wall and a lumen wall for each coronary artery identified in the medical images. In some embodiments, the system is then configured to determine the volume between the vessel wall and the lumen wall as plaque. In some embodiments, the system can be configured to identify plaque regions based on radiation density values ​​commonly associated with plaque, for example, by setting a predetermined threshold or range of radiation density values ​​commonly associated with plaque, with or without normalization using a normalization device.

[0276] Similarly, in some embodiments, the system can be configured to identify one or more fatty regions, such as epicardial fat, in medical images, e.g., using one or more AI and / or ML algorithms to automatically and / or dynamically identify one or more fatty regions. In some embodiments, the one or more AI and / or ML algorithms can be trained using a convolutional neural network (CNN) on a set of medical images in which fatty regions have been identified, thereby enabling the AI ​​and / or ML algorithm to automatically identify fatty regions directly from the medical images. In some embodiments, the system can be configured to identify fatty regions based on radiation density values ​​typically associated with fat, e.g., by setting a predetermined threshold or range of radiation density values ​​typically associated with fat, with or without normalization using a normalization device.

[0277] In some embodiments, the system may be configured to determine one or more vascular morphology and / or quantified plaque parameters at block 208. For example, in some embodiments, the system may be configured to determine the geometry and / or volume of the plaque and / or vascular region at block 201, the volume to surface area ratio or function of the plaque region at block 203, the heterogeneity or homogeneity index of the plaque region at block 205, the radiodensity of the plaque region and / or its composition by a range of radiodensity values ​​at block 207, the radiodensity to volume ratio of the plaque region at block 209, and / or the diffusivity of the plaque region at block 211. Additional details regarding the processes and techniques represented by blocks 208, 201, 203, 205, 207, 209, and 211 may be found in the discussion above with respect to FIG. 2A.

[0278] In some embodiments, at block 358, the system is configured to calculate or determine a numerical calculation or representation of coronary stenosis based on one or more vascular morphological parameters and / or quantified plaque parameters derived from medical images of the subject's coronary artery region. In some embodiments, the system may be configured to generate a weighted measure of the one or more vascular morphological parameters and / or quantified plaque parameters determined and / or derived from the raw medical images. For example, in some embodiments, the system may be configured to weight the one or more vascular morphological parameters and / or quantified plaque parameters equally. In some embodiments, the system may be configured to weight the one or more vascular morphological parameters and / or quantified plaque parameters differently. In some embodiments, the system may be configured to weight the one or more vascular morphological parameters and / or quantified plaque parameters logarithmically, algebraically, and / or utilizing another mathematical transformation. In some embodiments, the system is configured to calculate the stenosis at block 358 using the weighted measure and / or using only some of the vascular morphological parameters and / or quantified plaque parameters. In some embodiments, the system can be configured to calculate stenosis on a vessel-by-vessel or region-by-region basis.

[0279] In some embodiments, based on the calculated stenosis, the system is configured to determine a CAD-RADS score at block 360. This is in contrast to existing methods of determining CAD-RADS based on visual or general evaluation of medical images by a physician, which may yield non-reproducible results. However, in some embodiments described herein, the system can be configured to generate a reproducible and / or objective calculated CAD-RADS score based on automatic and / or dynamic image processing of raw medical images.

[0280] In some embodiments, at block 362, the system may be configured to determine the presence or risk of ischemia based on the calculated stenosis, one or more quantified plaque parameters and / or vascular morphology parameters derived from the medical images. For example, in some embodiments, the system may be configured to determine the presence or risk of ischemia by combining one or more of the above-mentioned parameters with or without weighting, or by using some or all of these parameters individually. In some embodiments, the system may be configured to determine the presence or risk of ischemia by comparing one or more of the calculated stenosis, one or more quantified plaque parameters and / or vascular morphology parameters to a database of known such parameters derived from medical images of other subjects, including, for example, healthy subjects and / or subjects at risk for cardiovascular events. In some embodiments, the system may be configured to calculate the presence or risk of ischemia per vessel or region.

[0281] In some embodiments, at block 364, the system may be configured to determine one or more quantified parameters of fat for one or more fatty regions identified from the medical image. For example, in some embodiments, the system may utilize any of the processes and / or techniques discussed herein with respect to deriving quantified parameters of plaque, such as those described in connection with blocks 208, 201, 203, 205, 207, 209, and 211. In particular, in some embodiments, the system may be configured to determine one or more parameters of fat including volume, geometry, radio density, and / or the like, of one or more fatty regions in the medical image.

[0282] In some embodiments, at block 366, the system may be configured to generate a risk assessment of a cardiovascular disease or event for the subject. In some embodiments, the generated risk assessment may include a risk score indicative of the subject's risk of coronary artery disease. In some embodiments, the system may generate the risk assessment based on one or more vascular morphology parameters, one or more quantified plaque parameters, one or more quantified fat parameters, calculated stenosis, ischemic risk, CAD-RADS score, and / or other analyses. In some embodiments, the system may be configured to generate a weighted measure of the subject's one or more vascular morphology parameters, one or more quantified plaque parameters, one or more quantified fat parameters, calculated stenosis, ischemic risk, and / or CAD-RADS score. For example, in some embodiments, the system may be configured to weight one or more of the above-mentioned parameters equally. In some embodiments, the system may be configured to weight one or more of these parameters differently. In some embodiments, the system may be configured to weight one or more of these parameters logarithmically, algebraically, and / or utilizing another mathematical transformation. In some embodiments, the system is configured to generate a coronary artery disease or cardiovascular event risk assessment for the subject at block 366 using the weighted criteria and / or using only some of these parameters.

[0283] In some embodiments, the system may be configured to generate a risk assessment of coronary artery disease or a cardiovascular event for the subject by combining one or more of the above-mentioned parameters, with or without weighting, or by using some or all of these parameters individually. In some embodiments, the system may be configured to generate a risk assessment of coronary artery disease or a cardiovascular event by comparing the subject's one or more vascular morphology parameters, one or more quantified plaque parameters, one or more quantified fat parameters, calculated stenosis, risk of ischemia, and / or CAD-RADS score to a database of known such parameters derived from medical images of other subjects, including, for example, healthy subjects and / or subjects at risk for a cardiovascular event.

[0284] Further, in some embodiments, the system can be configured to automatically and / or dynamically generate CAD-RADS modifiers based on one or more of the determined vascular morphological parameters, the set of quantified plaque parameters of one or more plaque areas, the quantified coronary stenosis, the determined presence or risk of ischemia, and / or the determined set of quantified lipid parameters. In particular, in some embodiments, the system can be configured to automatically and / or dynamically generate one or more applicable CAD-RADS modifiers for the subject, including, for example, one or more of non-diagnostic (N), stent (S), graft (G), or vulnerable (V), as defined by and used by CAD-RADS. For example, N can indicate that the exam is non-diagnostic, S can indicate the presence of a stent, G can indicate the presence of a coronary artery bypass graft, and V can indicate the presence of vulnerable plaque, e.g., indicating a low radiodensity value.

[0285] In some embodiments, the system can be configured to generate a suggested treatment for the subject based on the generated risk assessment of coronary artery disease, one or more vascular morphology parameters, one or more quantified plaque parameters, one or more quantified fat parameters, calculated stenosis, risk of ischemia, CAD-RADS score, and / or CAD-RADS modifiers derived from the raw medical images using image processing.

[0286] In some embodiments, one or more of the processes described herein in connection with Figure 3B can be repeated. For example, if medical images of the same subject are taken again at a later time, one or more of the processes described herein can be repeated and the analysis results can be used to track the quantified plaque, calculated stenosis, CAD-RADS score, and / or modifiers derived from the medical images, determined ischemic risk, quantified fat parameters, coronary artery disease risk assessment generated for the subject, and / or other purposes.

[0287] Disease Tracking In some embodiments, the systems, methods, and devices described herein can be configured to track the progression and / or disappearance of arterial and / or plaque-based diseases, such as coronary artery disease. For example, in some embodiments, the system can be configured to track the progression and / or disappearance of disease by automatically and / or dynamically analyzing multiple medical images obtained at different times using one or more techniques discussed herein and comparing different parameters derived therefrom. Thus, in some embodiments, the system can provide an automated disease tracking tool using non-invasive raw medical images as input that does not rely on subjective assessment.

[0288] In particular, in some embodiments, the system can be configured to utilize a four-category system for determining whether a subject is experiencing plaque stabilization or worsening. For example, in some embodiments, these categories can include: (1) "plaque progression" or "rapid plaque progression," (2) "mixed response-calcium dominant" or "non-rapid calcium dominant mixed response," (3) "mixed response-non-calcium dominant" or "non-rapid non-calcium dominant mixed response," or (4) "plaque disappearance."

[0289] In some embodiments, in the case of "plaque progression" or "rapid plaque progression", the total or relative volume of plaque increases. In some embodiments, in the case of "mixed reaction - calcium dominant" or "non-rapid calcium dominant mixed reaction", the plaque volume remains relatively constant or does not increase to the threshold level of "rapid plaque progression", but there is an overall progression of calcified plaque and an overall disappearance of non-calcified plaque. In some embodiments, in the case of "mixed reaction - non-calcium dominant" or "non-rapid non-calcium dominant mixed reaction", the plaque volume remains relatively constant or there is an overall progression of non-calcified plaque and an overall disappearance of calcified plaque. In some embodiments, in the case of "plaque disappearance", the total or relative volume of plaque decreases.

[0290] In some embodiments, these four categories can be expanded to be more granular, for example, including higher density versus lower density calcium plaque (e.g., >1000 Hounsfield units versus <1000 Hounsfield units), and / or more specifically categorize into a mixed response of predominantly calcium and predominantly non-calcified plaque. For example, in the case of a mixed response of predominantly non-calcified plaque, non-calcified plaque can further include necrotic core, fibro-fatty plaque, and / or fibrous plaque as separate categories within the overall umbrella of non-calcified plaque. Similarly, calcified plaque can be categorized as low density calcified plaque, medium density calcified plaque, and high density calcified plaque.

[0291] 3D is a flow chart illustrating an overview of an example embodiment of a method for disease tracking based on medical image analysis. For example, in some embodiments, a system can be configured to track the progression and / or resolution of a plaque-based disease or condition, such as coronary artery disease associated with or involving atherosclerosis, stenosis, ischemia, and / or the like, by analyzing one or more non-invasively obtained medical images.

[0292] As shown in FIG. 3D, in some embodiments, the system is configured to access, at block 372, a first set of plaque parameters derived from medical images of the subject at a first time point. In some embodiments, the medical images can be stored in the medical image database 100 and can include any of the types of medical images described above, including, for example, CT, non-contrast CT, contrast-enhanced CT, MR, DECT, spectral CT, and / or others. In some embodiments, the medical images of the subject can include the subject's coronary territories, coronary arteries, carotid arteries, renal arteries, abdominal aorta, cerebral arteries, lower extremities, and / or upper extremities. In some embodiments, the set of plaque parameters can be stored in a plaque parameter database 370, which can include any of the quantified plaque parameters described above in connection with blocks 208, 201, 203, 205, 207, 209, and / or 211.

[0293] In some embodiments, the system can be configured to directly access a first set of plaque parameters previously derived from medical images and / or stored in a plaque parameters database 370. In some embodiments, the plaque parameters database 370 can be locally accessible and / or remotely accessible by the system via a network connection. In some embodiments, the system can be configured to dynamically and / or automatically derive the first set of plaque parameters from medical images taken from a first time point.

[0294] In some embodiments, at block 374, the system may be configured to access a second medical image of the subject, which may be obtained from the subject at a later time than the medical images from which the first set of plaque parameters were derived. In some embodiments, the medical images may be stored in the medical image database 100 and may include any of the types of medical images described above, including, for example, CT, non-contrast CT, contrast-enhanced CT, MR, DECT, spectral CT, and / or others.

[0295] In some embodiments, at block 376, the system may be configured to dynamically and / or automatically derive a second set of plaque parameters from a second medical image taken from a second time point. In some embodiments, the second set of plaque parameters may include any of the quantified plaque parameters described above in connection with blocks 208, 201, 203, 205, 207, 209, and / or 211. In some embodiments, the system may be configured to store the derived or determined second set of plaque parameters in a plaque parameters database 370.

[0296] In some embodiments, at block 378, the system may be configured to analyze changes in one or more plaque parameters between a first set derived from medical images taken at a first time point and a second set derived from medical images taken at a later time point. For example, in some embodiments, the system may be configured to compare quantified plaque parameters between the two scans, such as, for example, radiodensity, volume, geometry, location, volume to surface area ratio or function, heterogeneity index, radiodensity composition, radiodensity composition as a function of volume, radiodensity to volume ratio, diffusivity, any combination or relationship thereof, and / or others, of one or more plaque regions. In some embodiments, the system may be configured to determine heterogeneity index of one or more plaque regions by generating a spatial mapping or three-dimensional histogram of radiodensity values ​​across the geometry of one or more plaque regions. In some embodiments, the system may be configured to analyze changes in one or more non-image-based metrics, such as, for example, serum biomarkers, genetics, omics, transcriptomics, microbiomics, and / or metabolomics.

[0297] In some embodiments, the system is configured to determine the change in plaque composition between two scans in terms of radiodensity or stable versus unstable plaque. For example, in some embodiments, the system is configured to determine the change between two scans in the percentage of high radiodensity or stable plaque versus low radiodensity or unstable plaque. In some embodiments, the system can be configured to track the change between two scans in high radiodensity plaque versus low radiodensity plaque. In some embodiments, the system can be configured to define high radiodensity plaque as having more than 1000 Hounsfield units and low radiodensity plaque as having less than 1000 Hounsfield units.

[0298] In some embodiments, at block 380, the system may be configured to determine plaque progression or disappearance, and / or any other related measurement, symptom, assessment, or associated disease, based on a comparison of one or more parameters derived from the two or more scans and / or changes in one or more non-image-based metrics, such as serum biomarkers, genetics, omics, transcriptomics, microbiomics, and / or metabolomics. For example, in some embodiments, the system may be configured to determine plaque progression and / or disappearance in general, atherosclerosis, stenosis, risk or presence of ischemia, and / or others. Further, in some embodiments, the system may be configured to automatically and / or dynamically generate a CAD-RADS score for the subject based on the quantified or calculated stenosis, as derived from the two medical images. Further details regarding the generation of a CAD-RADS score are described herein in connection with FIG. 3C. In some embodiments, the system may be configured to determine progression or disappearance in the subject's CAD-RADS score. In some embodiments, the system can be configured to compare the plaque parameters individually and / or in combination with one or more of them as a weighted measure. For example, in some embodiments, the system can be configured to weight the plaque parameters equally, differently, logarithmically, algebraically, and / or using another mathematical transformation. In some embodiments, the system can be configured to use only some or all of the quantified plaque parameters.

[0299] In some embodiments, the plaque progression state as determined by the system may include one of four categories including rapid plaque progression, non-rapid calcium dominated mixed response, non-rapid non-calcium dominated mixed response, or plaque disappearance. In some embodiments, the system is configured to classify the plaque progression state as rapid plaque progression if the subject's percentage atheroma volume increase is greater than 1% per year. In some embodiments, the system is configured to classify the plaque progression state as non-rapid calcium dominated mixed response if the subject's percentage atheroma volume increase is less than 1% per year and calcified plaque represents greater than 50% of the total new plaque formation. In some embodiments, the system is configured to classify the plaque progression state as non-rapid non-calcium dominated mixed response if the subject's percentage atheroma volume increase is less than 1% per year and non-calcified plaque represents greater than 50% of the total new plaque formation. In some embodiments, the system is configured to classify the plaque progression state as plaque disappearance if there is a decrease in the percentage of total atheroma volume.

[0300] In some embodiments, the system may be configured to generate a suggested treatment plan for the subject at block 382. For example, in some embodiments, the system may be configured to generate a suggested treatment plan for the subject based on the determined plaque progression or clearance, and / or any other relevant measurements, symptoms, assessments, or associated diseases, based on a comparison of one or more parameters derived from two or more scans.

[0301] In some embodiments, one or more of the processes described herein in connection with Figure 3D can be repeated, for example, one or more of the processes described herein can be repeated and the analytical results can be used for serial tracking of plaque-based disease and / or other purposes.

[0302] Determining the cause of calcium score changes In some embodiments, the systems, methods, and devices disclosed herein can be configured to generate analyses and / or reports that can determine the possible causes of an increase in calcium score. A high or increased calcium score alone does not represent any particular cause, either positive or negative. Instead, in general, various high or increased calcium scores can have various possible causes. For example, in some cases, a high or increased calcium score can be an indication of significant heart disease and / or an increased risk of a patient having a heart attack. In some cases, a high or increased calcium score can be an indication of an increased amount of exercise that a patient is performing (as exercise can convert fatty plaque in arterial blood vessels). In some cases, a high or increased calcium score can be an indication of a patient beginning a treatment regimen of statins that convert fatty plaque to calcium. Unfortunately, it is not possible to determine which of the above reasons are the possible causes of an increase in calcium score using blood tests alone. In some embodiments, by utilizing one or more techniques described herein, a system can be configured to determine the cause of a high or increased calcium score.

[0303] More specifically, in some embodiments, the system may be configured to track specific segments in the patient's arterial vessel wall in a manner that monitors the conversion of fatty material plaque lesions to near-calcified plaque deposits, which may aid in determining the cause of the calcium score increase, such as one or more of the causes identified above. Additionally, in some embodiments, the system may be configured to determine and / or use the location, size, shape, diffusivity, and / or attenuation radiodensity of one or more regions of calcified plaque to determine the cause of the calcium score increase. As a non-limiting example, if the density of calcium plaque increases, this may represent plaque stabilization due to treatment or lifestyle, but if new calcium plaque forms where it was not previously present (especially where the attenuation density is low), this may represent an adverse finding of disease progression rather than stabilization. In some embodiments, one or more of the processes and techniques described herein may be applied to non-contrast CT scans (such as ECG-gated coronary calcium score or non-ECG-gated chest CT), as well as contrast-enhanced CT scans (such as coronary CT angiograms).

[0304] As another non-limiting example, CT scan image acquisition parameters can be altered to improve understanding of calcium changes over time. As an example, conventional coronary calcium imaging is performed using a slice thickness of 2.5-3.0 mm, and detecting voxels / pixels of 130 Hounsfield units or greater. Alternative examples could be those that perform "thin" slice imaging, such as a slice thickness of 0.5 mm, as well as those that detect all Hounsfield unit densities below 130 but above a certain threshold (e.g., 100) that can identify low density calcium that may be missed by the arbitrary 130 Hounsfield unit threshold.

[0305] FIG. 3E is a flow chart showing an overview of one example embodiment of a method for determining the cause of a change in calcium score, whether an increase or decrease, based on medical image analysis.

[0306] As shown in FIG. 3E, in some embodiments, the system can be configured to access a first calcium score and / or a first set of plaque parameters for the subject at block 384. The first calcium score and / or the first set of plaque parameters can be derived from medical images of the subject and / or from blood tests at a first time point. In some embodiments, the medical images can be stored in medical image database 100 and can include any of the types of medical images described above, including, for example, CT, non-contrast CT, contrast-enhanced CT, MR, DECT, spectral CT, and / or others. In some embodiments, the medical images of the subject can include the coronary artery territories, coronary arteries, carotid arteries, renal arteries, abdominal aorta, cerebral arteries, lower extremities, and / or upper extremities of the subject. In some embodiments, the set of plaque parameters can be stored in plaque parameter database 370, which can include any of the quantified plaque parameters described above in connection with blocks 208, 201, 203, 205, 207, 209, and / or 211.

[0307] In some embodiments, the system may be configured to directly access and / or retrieve the first calcium score and / or first set of plaque parameters stored in the calcium score database 398 and / or the plaque parameter database 370, respectively. In some embodiments, the plaque parameter database 370 and / or the calcium score database 298 may be locally accessible and / or remotely accessible by the system via a network connection. In some embodiments, the system may be configured to dynamically and / or automatically derive the first set of plaque parameters and / or calcium score from medical images and / or blood tests of the subject taken from the first time point.

[0308] In some embodiments, at block 386, the system may be configured to access a second calcium score and / or a second medical image of the subject, which may be obtained from the subject at a later time than the first calcium score and / or medical image from which the first set of plaque parameters were derived. For example, in some embodiments, the second calcium score may be derived from a second medical image and / or a second blood test taken from the subject at a second time. In some embodiments, the second calcium score may be stored in the calcium score database 398. In some embodiments, the medical images may be stored in the medical image database 100 and may include any of the types of medical images described above, including, for example, CT, non-contrast CT, contrast-enhanced CT, MR, DECT, spectral CT, and / or others.

[0309] In some embodiments, at block 388, the system may be configured to compare the first calcium score to the second calcium score to determine a change in calcium score. However, as discussed above, this alone generally does not provide insight into the cause of the change in calcium score, if any. In some embodiments, if there is no statistically significant calcium score change between the two readings, e.g., if the difference is below a predetermined threshold, the system may be configured to terminate the analysis of the change in calcium score. In some embodiments, if there is a statistically significant calcium score change between the two readings, e.g., if the difference is above a predetermined threshold, the system may be configured to continue its analysis.

[0310] In particular, in some embodiments, at block 390, the system may be configured to dynamically and / or automatically derive a second set of plaque parameters from a second medical image taken from a second time point. In some embodiments, the second set of plaque parameters may include any of the quantified plaque parameters described above in connection with blocks 208, 201, 203, 205, 207, 209, and / or 211. In some embodiments, the system may be configured to store the derived or determined second set of plaque parameters in a plaque parameters database 370.

[0311] In some embodiments, at block 392, the system may be configured to analyze changes in one or more plaque parameters between a first set derived from medical images taken at a first time point and a second set derived from medical images taken at a later time point. For example, in some embodiments, the system may be configured to compare quantified plaque parameters between the two scans, such as, for example, radiodensity, volume, geometry, location, volume to surface area ratio or function, heterogeneity index, radiodensity composition, radiodensity composition as a function of volume, radiodensity to volume ratio, diffusivity, any combination or relationship thereof, and / or the like, of one or more plaque regions and / or one or more regions surrounding the plaque. In some embodiments, the system may be configured to determine a heterogeneity index of one or more plaque regions by generating a spatial mapping or three-dimensional histogram of radiodensity values ​​across the geometry of one or more plaque regions. In some embodiments, the system is configured to analyze changes in one or more non-image-based metrics, such as, for example, serum biomarkers, genetics, omics, transcriptomics, microbiomics, and / or metabolomics.

[0312] In some embodiments, the system is configured to determine the change in plaque composition between two scans in terms of radiodensity or stable versus unstable plaque. For example, in some embodiments, the system is configured to determine the change between two scans in the percentage of high radiodensity or stable plaque versus low radiodensity or unstable plaque. In some embodiments, the system can be configured to track the change between two scans in high radiodensity plaque versus low radiodensity plaque. In some embodiments, the system can be configured to define high radiodensity plaque as having more than 1000 Hounsfield units and low radiodensity plaque as having less than 1000 Hounsfield units.

[0313] In some embodiments, the system can be configured to compare the plaque parameters individually and / or in combination with one or more of them as a weighted measure. For example, in some embodiments, the system can be configured to weight the plaque parameters equally, differently, logarithmically, algebraically, and / or using another mathematical transformation. In some embodiments, the system can be configured to use only some or all of the quantified plaque parameters.

[0314] In some embodiments, at block 394, the system may be configured to characterize a change in calcium score for the subject based on a comparison of one or more plaque parameters, whether individually and / or in combination or weighted. In some embodiments, the system may be configured to characterize a change in calcium score as positive, neutral, or negative. For example, in some embodiments, if a comparison of one or more plaque parameters reveals that, for the subject as a whole, plaque is stabilizing or exhibiting high radiodensity values ​​without generating any new plaque, the system may report a change in calcium score as positive. In contrast, if a comparison of one or more plaque parameters reveals that, for the subject as a whole, plaque is destabilizing, for example, by generation of new unstable plaque regions having low radiodensity values, without generating any new plaque, the system may report a change in calcium score as negative. In some embodiments, the system may be configured to utilize any or all of the techniques for plaque quantification and / or plaque-based disease analysis tracking discussed herein, including those discussed in connection with Figures 3A, 3B, 3C, and 3D.

[0315] As a non-limiting example, in some embodiments, the system can be configured to characterize the cause of the calcium score change based on determining and comparing the change in the ratio between the volume and radiodensity of one or more plaque regions between the two scans. Similarly, in some embodiments, the system can be configured to characterize the cause of the calcium score change based on determining and comparing the change in the diffusivity and / or radiodensity of one or more plaque regions between the two scans. For example, if the radiodensity of the plaque region is increasing, the system can be configured to characterize the calcium score change or increase as positive. In some embodiments, if the system identifies one or more new plaque regions in the second image that were not present in the first image, the system can be configured to characterize the calcium score change as negative. In some embodiments, if the system determines that the volume to surface area ratio of one or more plaque regions is decreasing between the two scans, the system can be configured to characterize the calcium score change as positive. In some embodiments, if the system determines, for example, by generating and / or analyzing a spatial mapping of radiodensity values, that the heterogeneity or heterogeneity index of the plaque region has decreased between the two scans, the system can be configured to characterize the calcium score change as being positive.

[0316] In some embodiments, the system is configured to utilize AI, ML, and / or other algorithms to characterize calcium score changes based on one or more plaque parameters derived from medical images. For example, in some embodiments, the system can be configured to utilize AI and / or ML algorithms that are trained using CNNs and / or a known dataset of medical images in which plaque parameters to be combined with calcium scores have been identified. In some embodiments, the system can be configured to characterize calcium score changes by accessing a known dataset of calcium scores stored in a database. For example, the known dataset can include calcium score changes in other subjects in the past, and / or medical images, and / or a dataset of plaque parameters derived therefrom. In some embodiments, the system can be configured to characterize calcium score changes and / or determine the cause of the changes on a vessel-by-vessel, segment-by-segment, plaque-by-plaque, and / or subject-by-subject basis.

[0317] In some embodiments, the system can be configured to generate a suggested treatment plan for the subject at block 396. For example, in some embodiments, the system can be configured to generate a suggested treatment plan for the subject based on the change in and / or characterization of the subject's calcium score.

[0318] In some embodiments, one or more of the processes described herein in connection with Figure 3E can be repeated. For example, one or more of the processes described herein can be repeated and the results of the analysis can be used to continuously track and / or characterize changes in the calcium score of the subject, and / or for other purposes.

[0319] Prognosis of cardiovascular events In some embodiments, the systems, devices, and methods described herein are configured to generate a prognosis of a cardiovascular event in a subject based on one or more of the medical image-based analysis techniques described herein. For example, in some embodiments, the system is configured to determine whether a patient is at risk for a cardiovascular event based on the amount of malignant plaque accumulation in the patient's arterial blood vessels. For this purpose, a cardiovascular event can include clinical major cardiovascular events, such as heart attack, stroke, or death, as well as disease progression and / or ischemia.

[0320] In some embodiments, the system can identify risk of a cardiovascular event based on the ratio of the amount and / or volume of malignant plaque accumulation to the total surface area and / or volume of some or all of the patient's arterial blood vessels. In some embodiments, if the ratio exceeds a certain threshold, the system can be configured to output a certain risk factor and / or number and / or level associated with the patient. In some embodiments, the system is configured to determine whether the patient is at risk of a cardiovascular event based on the absolute amount or volume, or the percentage of the amount or volume, of malignant plaque accumulation in the patient's arterial blood vessels compared to the total volume of some or all of the arterial blood vessels. In some embodiments, the system is configured to determine whether the patient is at risk of a cardiovascular event based on results from the patient's blood chemistry or biomarker tests, for example, whether the patient's particular blood chemistry or biomarker tests exceed a certain threshold level. In some embodiments, the system is configured to receive as input from a user or other system and / or access the patient's blood chemistry or biomarker test data from a database system. In some embodiments, the system can be configured to utilize inputs from other imaging data related to the non-coronary cardiovascular system, such as plaque, vessel morphology, and / or stenosis-related arterial information, as well as opposing left ventricular mass, cavity volume and size, valve morphology, vessel (e.g., aorta, pulmonary artery) morphology, fat, and / or lung and / or bone health. In some embodiments, the system can utilize the output risk factors to generate a treatment plan suggestion. For example, the system can be configured to output a treatment plan that involves administration of cholesterol-lowering drugs, such as statins, to convert soft malignant plaque to hard plaque that is safer and more stable for the patient. In general, substantially calcified hard plaques may be at significantly lower risk of rupturing at the interface with the arterial vessel, thereby reducing the probability of blood clots forming within the arterial vessel, thereby reducing the patient's risk of heart attack or other cardiac events.

[0321] FIG. 4A is a flow chart illustrating an overview of one example embodiment of a method for prognosticating a cardiovascular event based on and / or derived from medical image analysis.

[0322] As shown in FIG. 4A, in some embodiments, the system may be configured to access medical images, such as CT scans of a coronary artery region of a subject, which may be stored in the medical image database 100, at block 202. Further, in some embodiments, the system may be configured to identify one or more arteries at block 204 and / or one or more plaque regions at block 206. Additionally, in some embodiments, the system may be configured to determine one or more vessel morphology and / or quantified plaque parameters at block 208. For example, in some embodiments, the system may be configured to determine the geometry and / or volume of the plaque and / or the region of the vessel, a volume-to-surface area ratio or function of the region of the plaque, a heterogeneity or homogeneity index of the region of the plaque, a radiodensity of the region of the plaque and / or its composition by a range of radiodensity values, a radiodensity-to-volume ratio of the region of the plaque, and / or a diffusivity of the region of the plaque. Additionally, in some embodiments, at block 210, the system may be configured to classify one or more plaque regions as stable versus unstable or good versus bad based on one or more vascular morphology parameters and / or quantified plaque parameters determined and / or derived from the raw medical images. Additional details regarding the processes and techniques represented by blocks 202, 204, 206, 208, and 210 may be found in the discussion above with respect to FIG. 2A.

[0323] In some embodiments, the system is configured to generate a ratio of malignant plaques to blood vessels in which malignant plaques appear at block 412. More specifically, in some embodiments, the system may be configured to determine the total surface area of ​​the blood vessels identified on the medical image and the surface area of ​​all areas of malignant or unstable plaques in the blood vessel. Based on the above, in some embodiments, the system may be configured to generate a ratio of the surface area of ​​all malignant plaques in a particular blood vessel and the surface area of ​​the entire blood vessel or a portion thereof shown on the medical image. Similarly, in some embodiments, the system may be configured to determine the total volume of the blood vessels identified on the medical image and the volume of all areas of malignant or unstable plaques in the blood vessel. Based on the above, in some embodiments, the system may be configured to generate a ratio of the volume of all malignant plaques in a particular blood vessel and the volume of the entire blood vessel or a portion thereof shown on the medical image.

[0324] In some embodiments, at block 414, the system is further configured to determine the absolute total volume and / or surface area of ​​all malignant or unstable plaque identified in the medical image. Also, in some embodiments, at block 416, the system is configured to determine the absolute total volume of all plaque, including benign and malignant plaque, identified in the medical image. Also, in some embodiments, at block 418, the system may be configured to access or obtain results from the patient's blood chemistry and / or biomarker testing, and / or other non-imaging test results. Also, in some embodiments, at block 422, the system may be configured to access and / or analyze one or more non-coronary cardiovascular medical images.

[0325] In some embodiments, at block 420, the system may be configured to analyze one or more of the generated ratios of malignant plaque to vessels, whether from surface area or volume, absolute total volume of malignant plaque, absolute total volume of plaque, blood chemistry and / or biomarker test results, and / or from analysis of one or more non-coronary cardiovascular medical images, to determine whether one or more of these parameters, either individually and / or in combination, exceed a predefined threshold. For example, in some embodiments, the system may be configured to analyze one or more of the above-mentioned parameters individually by comparing them to one or more reference values ​​for healthy subjects and / or subjects at risk for cardiovascular events. In some embodiments, the system may be configured to analyze a combination, such as a weighted criteria, of one or more of the above-mentioned parameters by comparing the combined or weighted criteria to one or more reference values ​​for healthy subjects and / or subjects at risk for cardiovascular events. In some embodiments, the system may be configured to weight one or more of these parameters equally. In some embodiments, the system may be configured to weight one or more of these parameters differently. In some embodiments, the system may be configured to weight one or more of these parameters logarithmically, algebraically, and / or utilizing another mathematical transformation. In some embodiments, the system may be configured to utilize only some of the above-mentioned parameters individually, in combination, and / or as part of the weighted criteria.

[0326] In some embodiments, at block 424, the system is configured to generate a prognosis of a cardiovascular event for the subject. In particular, in some embodiments, the system is configured to generate a prognosis of a cardiovascular event based on one or more of the results of an analysis of a generated ratio of malignant plaque to vessel, whether by surface area or volume, absolute total volume of malignant plaque, absolute total volume of plaque, blood chemistry and / or biomarker test results, and / or the results of an analysis of one or more non-coronary cardiovascular medical images. In some embodiments, the system is configured to utilize AI, ML, and / or other algorithms to generate the prognosis. In some embodiments, the generated prognosis includes a risk score or risk assessment of a cardiovascular event for the subject. In some embodiments, the cardiovascular event can include one or more of atherosclerosis, stenosis, ischemia, heart attack, and / or others.

[0327] In some embodiments, at block 426, the system can be configured to generate a suggested treatment plan for the subject. For example, in some embodiments, the system can be configured to generate a suggested treatment plan for the subject based on the change in and / or characterization of the subject's calcium score. In some embodiments, the generated treatment plan can include the use of statins, lifestyle changes, and / or surgery.

[0328] In some embodiments, one or more of the processes described herein in relation to Figure 4A can be repeated. For example, one or more of the processes described herein can be repeated, and the analysis results can be used for the ongoing prognosis of a cardiovascular event in the subject and / or for other purposes.

[0329] Patient-specific stent determination In some embodiments, the systems, methods, and devices described herein can be used to determine and / or generate one or more parameters for patient-specific stent and / or implantation selection or guidance. In particular, in some embodiments, the systems disclosed herein can be used to dynamically and automatically determine the type, length, diameter, gauge, strength, and / or any other stent parameters required for a particular patient based on processing of medical imaging data, e.g., using AI, ML, and / or other algorithms.

[0330] In some embodiments, by determining one or more patient-specific stent parameters that are optimal for a particular arterial area, the system may reduce a patient's risk of complications and / or insurance risk, since if a stent that is too large is implanted, the arterial wall may stretch and thin too much, resulting in potential rupture, or undesirably high blood flow, or other problems, whereas if a stent that is too small is implanted, the arterial wall may not advance and open sufficiently, resulting in too little blood flow, or other problems.

[0331] In some embodiments, the system is configured to dynamically identify an area of ​​stenosis in an artery, dynamically determine an appropriate diameter for the identified area of ​​the artery, and / or automatically select a stent from a selection of multiple available stents. In some embodiments, the selected stent can be configured to leave the arterial area open to the determined appropriate arterial diameter after implantation. In some embodiments, the appropriate arterial diameter is determined to be equal or substantially equal to the diameter that would be the natural state in the absence of stenosis. In some embodiments, the system can be configured to dynamically generate a patient-specific surgical plan to implant the selected stent in the identified arterial area. For example, the system can be configured to determine whether an arterial bifurcation is near the identified arterial area and generate a patient-specific surgical plan to insert two guidewires to address the bifurcation and / or determine a location to constrain and insert a second stent into the bifurcation.

[0332] FIG. 4B is a flow chart illustrating an overview of one example embodiment of a method for determining patient-specific stent parameters based on medical image analysis.

[0333] As shown in FIG. 4B, in some embodiments, the system may be configured to access medical images, such as a CT scan, of a coronary artery region of a subject at block 202. Further, in some embodiments, the system may be configured to identify one or more arteries at block 204 and / or one or more plaque regions at block 206. Additionally, in some embodiments, the system may be configured to determine one or more vessel morphology and / or quantified plaque parameters at block 208. For example, in some embodiments, the system may be configured to determine the geometry and / or volume of the plaque and / or vessel region at block 201, a volume to surface area ratio or function of the region of plaque at block 203, a heterogeneity or homogeneity index of the region of plaque at block 205, a radiodensity of the region of plaque and / or its composition by a range of radiodensity values ​​at block 207, a radiodensity to volume ratio of the region of plaque at block 209, and / or a diffusivity of the region of plaque at block 211. Additional details regarding the processes and techniques represented by blocks 202, 204, 206, 208, 201, 203, 205, 207, 209, and 211 can be found in the discussion above regarding FIG. 2A.

[0334] In some embodiments, at block 440, the system may be configured to analyze the medical image to determine one or more vascular parameters, such as diameter, curvature, vessel morphology, vessel wall, lumen wall, and / or the like. In some embodiments, the system may be configured to determine or derive one or more vascular parameters as shown in the medical image, e.g., with a stenosis in a particular region along the vessel. In some embodiments, the system may be configured to determine one or more vascular parameters without the stenosis. For example, in some embodiments, the system may be configured to graphically and / or hypothetically remove the stenosis or plaque from the vessel to determine the diameter, curvature, and / or the like of the vessel if the stenosis was not present.

[0335] In some embodiments, at block 442, the system may be configured to determine whether a stent is recommended for the subject and, if so, determine one or more recommended parameters of the patient-specific stent based on the medical analysis. For example, in some embodiments, the system may be configured to analyze one or more of the identified vascular morphology parameters, the quantified plaque parameters, and / or the vascular parameters. In some embodiments, the system may be configured to utilize AI, ML, and / or other algorithms. In some embodiments, the system may be configured to analyze one or more of the above-mentioned parameters individually, in combination, and / or as a weighted criteria. In some embodiments, one or more of these parameters derived from the medical images may be compared, either individually or in combination, to one or more reference values ​​derived or collected from other subjects, including subjects who have and have not been implanted with stents. In some embodiments, based on the determined parameters of the patient-specific stent, the system may be configured to determine the selection of an existing stent that matches those parameters and / or generate manufacturing instructions for manufacturing the patient-specific stent having the stent parameters derived from the medical images. In some embodiments, the system can be configured to recommend a stent diameter that is less than or substantially equal to the diameter of the artery in the absence of a stenosis.

[0336] In some embodiments, at block 444, the system may be configured to generate a recommended surgical plan for stent implantation based on the analyzed medical images. For example, in some embodiments, the system may be configured to determine whether bifurcations are present and / or generate guidelines for positioning a guidewire and / or a stent relative to the patient pre-operatively based on the medical images. Thus, in some embodiments, the system may be configured to generate a detailed surgical plan that is specific to a particular patient based on medical image analysis of the plaque and / or other parameters.

[0337] In some embodiments, at block 446, the system is configured to access or obtain one or more medical images after stent implantation. In some embodiments, at block 448, the system can be configured to analyze the accessed medical images to perform a post-implant analysis. For example, in some embodiments, the system can be configured to derive one or more vascular morphology and / or plaque parameters after stent implantation, including any of those discussed herein in connection with block 208. Based on the above analysis, in some embodiments, the system can further generate a suggested treatment in some embodiments, such as, for example, a recommendation for use of a statin or other drug, a lifestyle change, further surgery or stent implantation, and / or the like.

[0338] In some embodiments, one or more of the processes described herein in connection with Figure 4B can be repeated. For example, one or more of the processes described herein can be repeated and the analysis results can be used to determine additional patient-specific stent needs and / or parameters for the patient, and / or for other purposes.

[0339] Patient-specific reporting In some embodiments, the system is configured to dynamically generate a patient-specific report based on an analysis of the processed data generated from the raw CT scan data. In some embodiments, the patient-specific report is dynamically generated based on the processed data. In some embodiments, the report is dynamically generated based on a selection and / or combination of specific phrases from the database, where the specific words, terms, and / or phrases are modified to be specific to the patient and the patient's identified medical problem. In some embodiments, the system is configured to dynamically select one or more images from the image scanning data and / or the system-generated image views described herein, where the selected one or more images are dynamically inserted into the report to generate the patient-specific report based on an analysis of the processed data.

[0340] In some embodiments, the system is configured to dynamically annotate one or more selected images for insertion into a patient-specific report, where the annotations are patient-specific and / or based on data processing performed by the devices, methods, and systems disclosed herein, for example, annotating one or more images to include markings or other indicators indicating where along an artery significant malignant plaque buildup is present.

[0341] In some embodiments, the system is configured to dynamically generate reports based on past and / or current medical data. For example, in some embodiments, the system can be configured to show how a patient's cardiovascular health has changed over a period of time. In some embodiments, the system is configured to dynamically generate phrases and / or select phrases from a database to specifically describe the patient's cardiovascular health and / or how cardiovascular disease has changed in the patient's body.

[0342] In some embodiments, the system is configured to dynamically select one or more medical images from a previous and / or current medical scanning for insertion into the medical report to show how cardiovascular disease has changed over time in the patient's body, e.g., by showing past and current images juxtaposed against each other, or by showing a past image superimposed on a current image, allowing a user to move or fade or toggle between past and current images.

[0343] In some embodiments, the patient-specific report is an interactive report that allows the user to interact with certain images, videos, animations, augmented reality (AR), virtual reality (VR), and / or features of the report. In some embodiments, the system is configured to insert dynamically generated illustrations or images of the patient's arterial blood vessels into the patient-specific report to highlight certain blood vessels and / or portions of blood vessels that contain or may contain vascular disease requiring examination or further analysis. In some embodiments, the dynamically generated patient-specific report is configured to show the blood vessel walls to the user using AR and / or VR.

[0344] In some embodiments, the system is configured to insert any ratios and / or dynamically generated data into a dynamically generated report using the methods, systems, and devices disclosed herein. In some embodiments, the dynamically generated report includes a radiology report. In some embodiments, the dynamically generated report is in the form of an editable document, such as Microsoft Word®, to allow a physician to edit the report. In some embodiments, the dynamically generated report is stored in a PACS (Picture Archiving and Communication System) or other EMR (Electronic Medical Record) system.

[0345] In some embodiments, the system is configured to convert and / or translate data from imaging into infographics in pictorial or video format, with or without audio, to accurately convey information in a form that is better understandable to any patient to improve literacy. In some embodiments, this method of improving literacy is coupled to a severity classification tool that defines lower risk at higher literacy and higher risk at lower literacy. In some embodiments, the output of these reports may be patient derived and / or patient specific. In some embodiments, actual patient imaging data (e.g., from a CT of the patient) may be coupled to graphics from the CT of the patient and / or drawings from the CT to further explain the findings. In some embodiments, actual patient imaging data, graphics data, and / or drawing data may be coupled to explanatory graphics that are not from the patient but may help the patient to understand better (e.g., a video about lipid-rich plaque).

[0346] In some embodiments, these patient reports can be imported into an application that allows for tracking of disease over time in relation to controlling cardiovascular risk factors such as diabetes or hypertension. In some embodiments, the app and / or user interface allows for tracking of blood glucose and blood pressure over time and / or can correlate changes in images over time in a manner that enhances risk prediction.

[0347] In some embodiments, the system can be configured to generate a patient-specific video report based on the processed data generated from the raw CT data. In some embodiments, the system is configured to generate and / or provide a personalized video viewing experience for the user that can be programmed to automatically and dynamically change content based on imaging findings, associated automated computational diagnostics, and / or prognostic algorithms. In some embodiments, the viewing method is through a video experience that can be in the form of a regular 2D video, as opposed to a traditional report, and / or through a mixed reality video experience through AR or VR. In some embodiments, in both 2D and mixed reality cases, the personalized video experience can interact with the patient to predict the risk of heart attack, disease progression rate, and / or prognosis of the patient, such as ischemia.

[0348] In some embodiments, the system can be configured to dynamically generate a video report that includes both cartoon images and / or animations along with audio content combined with actual CT image data from the patient. In some embodiments, the dynamically generated video medical report is dynamically narrated based on selected phrases, terms, and / or other content from a database such that a voice synthesizer or pre-made audio content can be used for playback during the video report. In some embodiments, the dynamically generated video medical report is configured to include any of the images disclosed herein. In some embodiments, the dynamically generated video medical report can be configured to dynamically select one or more medical images to insert into the video medical report from previous and / or current medical scanning to show how cardiovascular disease has changed over time in the patient's body. For example, in some embodiments, the report can show past and current images juxtaposed next to each other. In some embodiments, the report can show past images superimposed on the current image, thereby allowing the user to toggle or move or fade between past and current images. In some embodiments, the dynamically generated video medical report can be configured to show an actual medical image, such as a CT medical image, in the image report and then transition to an illustrated or cartooned version (partially or fully illustrated or cartooned) of the actual medical image, thereby highlighting particular features of the patient's arteries. In some embodiments, the dynamically generated video medical report is configured to show the blood vessel walls to the user using AR and / or VR.

[0349] 5A is a flow chart showing an overview of an example embodiment of a method for generating a medical report regarding patient characteristics based on medical image analysis. As shown in FIG. 5A, in some embodiments, the system can be configured to access medical images at block 202. In some embodiments, the medical images can be stored in a medical image database 100. Additional details regarding the types of medical images and other processes and techniques represented in block 202 can be found in the description above regarding FIG. 2A.

[0350] In some embodiments, the system is configured to identify one or more arteries, plaque, and / or fat in the medical image, for example using AI, ML, and / or other algorithms, at block 354. Additional details regarding the types of medical images and other processes and techniques represented in block 354 can be found in the discussion above with respect to FIG.

[0351] In some embodiments, the system may be configured to determine one or more vascular morphology and / or quantified plaque parameters at block 208. For example, in some embodiments, the system may be configured to determine the geometry and / or volume of the plaque and / or vascular region at block 201, the volume to surface area ratio or function of the region of plaque at block 203, the heterogeneity or homogeneity index of the region of plaque at block 205, the radiodensity of the region of plaque and / or its composition by a range of radiodensity values ​​at block 207, the radiodensity to volume ratio of the region of plaque at block 209, and / or the diffusivity of the region of plaque at block 211. Additional details regarding the processes and techniques represented by blocks 208, 201, 203, 205, 207, 209, and 211 may be found in the discussion above with respect to FIG. 2A.

[0352] In some embodiments, at block 508, the system can be configured to determine and / or quantify stenosis, atherosclerosis, risk of ischemia, risk of a cardiovascular event or disease, and / or the like. The system can be configured to utilize any of the techniques and / or algorithms described herein, including but not limited to those described above in connection with blocks 358 and 366 of FIG.

[0353] In some embodiments, the system may be configured to generate an annotated medical image and / or a quantized color map using the analysis results derived from the medical image at block 510. For example, in some embodiments, the system may be configured to generate a quantized map indicative of one or more arteries, plaque, fat, benign plaque, malignant plaque, vascular morphology, and / or the like.

[0354] In some embodiments, at block 512, the system may be configured to determine the patient's plaque and / or disease progression, for example, based on analysis of previously obtained medical images of the subject. In some embodiments, the system may be configured to utilize any of the algorithms or techniques described herein related to disease tracking, including but not limited to those generally described in connection with block 380 and / or FIG.

[0355] In some embodiments, the system may be configured to generate a suggested treatment plan for the patient based on the determined progression of the plaque and / or disease at block 514. In some embodiments, the system may be configured to utilize any of the algorithms or techniques described herein relating to disease tracking and treatment generation, including but not limited to those generally described in connection with block 382 and / or FIG.

[0356] In some embodiments, at block 516, the system may be configured to generate a patient-specific report. The patient-specific report may include one or more medical images of the patient and / or derived graphics thereof. For example, in some embodiments, the patient report may include one or more annotated medical images and / or quantified color maps. In some embodiments, the patient-specific report may include one or more vascular morphology and / or quantified plaque parameters derived from the medical images. In some embodiments, the patient-specific report may include quantified stenosis, atherosclerosis, ischemia, risk of cardiovascular events or disease, CAD-RADS score, and / or progression or tracking of any of the above. In some embodiments, the patient-specific report may include suggested treatments, such as statins, lifestyle changes, and / or surgery.

[0357] In some embodiments, the system may be configured to access and / or retrieve from the patient report database 500 one or more phrases, characterizations, graphics, videos, audio files, and / or the like that are applicable and / or usable to generate a patient-specific report. In generating a patient-specific report, in some embodiments, the system may be configured to compare one or more parameters, such as those described above and / or derived from the patient's medical images, to one or more parameters previously derived from other patients. For example, in some embodiments, the system may be configured to compare one or more quantified plaque parameters derived from the patient's medical images to one or more quantified plaque parameters derived from the medical images of other patients of a similar or same age group. Based on the comparison, in some embodiments, the system may be configured to determine which phrases, characterizations, graphics, videos, audio files, and / or the like to include in the patient-specific report, for example, by identifying similar previous cases. In some embodiments, the system may be configured to utilize AI and / or ML algorithms to generate the patient-specific report. In some embodiments, the patient-specific report may include written, AR experience, VR experience, video, and / or audio components.

[0358] 5B-5I illustrate an example embodiment of a medical report regarding patient characteristics generated based on medical image analysis. In particular, FIG. 5B illustrates the cover page of an example patient-specific report.

[0359] 5C-5I show portions of an example patient-specific report. In some embodiments, the patient-specific report generated by the system may include only some or all of these illustrated portions. As shown in FIG. 5C-5I, in some embodiments, the patient-specific report includes visualization of one or more arteries and / or portions thereof, such as, for example, the right coronary artery (RCA), the right posterior descending artery (R-PDA), the right posterior lateral branch (R-PLB), the left large (LM) artery and the left anterior descending (LAD) artery, the first diagonal (D1) artery, the second diagonal (D2) artery, the circumflex (Cx) artery, the first obtuse marginal branch (OM1), the second obtuse marginal branch (OM2), the intermediate branch (RI), and / or others. In some embodiments, for each artery included in the report, the system is configured to generate a linear image of a simple trace along the length of the vessel, such as, for example, at the proximal, intermediate, and / or distal portions of the artery.

[0360] In some embodiments, the patient-specific report generated by the system includes quantified measures of various plaque and / or vascular morphology related parameters present in the blood vessel. In some embodiments, for each or a portion of the arteries included in the report, the system is configured to include in the patient-specific report quantified measures of total plaque volume, total volume of low density or non-calcified plaque, total value of non-calcified plaque, and / or total volume of calcified plaque, generated and / or derived from medical images of the patient. Furthermore, in some embodiments, for each or a portion of the arteries included in the report, the system is configured to include in the patient-specific report quantified measures of stenosis severity, such as, for example, the percentage of maximum diameter stenosis in the artery, generated and / or derived from medical images of the patient. In some embodiments, for each or a portion of the arteries included in the patient-specific report, the system is configured to include in the patient-specific report quantified measures of vascular remodeling, such as, for example, the maximum remodeling index, generated and / or derived from medical images of the patient.

[0361] Visualization / GUI Atherosclerosis, a buildup of fatty deposits, cholesterol and other substances (e.g., plaque) in and on the walls of arteries that can restrict blood flow. The plaque can burst and induce a blood clot. Although atherosclerosis is often considered a heart problem, it can affect arteries anywhere in the body. However, determining information about plaque in coronary arteries can be difficult, in part, due to incomplete imaging data, aberrations that may be present in coronary artery images (e.g., due to patient motion), and differences in the manifestation of plaque in different patients. Thus, neither calculated information derived from CT images, nor visual inspection of CT images alone can provide sufficient information to determine conditions present in a patient's coronary arteries. Portions of this disclosure describe information that can be determined from CT images using automated or semi-automated processes. For example, a machine learning process that has been trained on thousands of CT scans can be used to determine information depicted in the CT images, and / or an analyst can be utilized to inspect and improve the results of the machine learning process, and the example user interface described herein can provide the determined information to another analyst or physician. While information determined from CT images is highly beneficial in accessing the condition of a patient's coronary arteries, visual analysis of the coronary arteries by a skilled physician, in conjunction with information determined from CT images at hand, allows for a more comprehensive assessment of a patient's coronary arteries. As presented herein, an embodiment of the system facilitates analysis and visualization of the vessel lumen, vessel wall, plaque, and stenosis in and around the coronary vessels. The system can display the vessels in multiplanar formats, cross-sectional views, 3D views of the coronary artery tree, axial, sagittal, and coronal views based on a set of computed tomography (CT) images, for example, generated by a CT scan of the patient's blood vessels. The CT images can be Digital Imaging and Communications in Medicine (DICOM) images, a standard for communication and management of medical imaging information and related data. CT image, or CT scan, as used herein, is a broad term that refers to a picture of structures within the body created by a computer-controlled scanner.For example, by a scanner using an X-ray beam. However, it is recognized that other radiation sources and / or imaging systems may create a set of CT-like images. Thus, use of the term "CT image" herein may refer to any type of imaging system with any type of imaging source that creates a set of images depicting "slices" of structures within the body, unless otherwise indicated. One key aspect of the user interface described herein is the precise correlation of displayed images and information of the CT images. Locations in the CT images displayed on portions (i.e., "panels") of the user interface are precisely correlated by the system such that the same locations are simultaneously displayed in different views. By simultaneously displaying portions of a coronary vessel in, for example, two, three, four, five, or six views, allowing the practitioner to investigate a particular location of the coronary vessel in one view and have the other two to six views correspondingly show the exact same location, a significant amount of insight into the condition of the vessel is provided, allowing the practitioner / analyst to quickly and easily visually integrate the presented information to obtain a comprehensive and accurate understanding of the condition of the coronary vessel being examined.

[0362] Advantageously, the present disclosure allows CT images and data to be analyzed in a more useful and accurate manner, and allows users to interact with, analyze images and data in a more analytically useful manner, and / or perform computational analysis in a more useful manner, for example, to detect conditions requiring attention. The graphical user interface allows users in the processes described herein to perform visualizations that would otherwise be difficult to define relationships between different information and views of the coronary artery. In one example, simultaneously displaying a portion of a coronary artery in CMPR, SMPR, and cross-sectional views can provide an analyst with insight into plaque or stenosis associated with the coronary artery that may not otherwise be perceptible using fewer views. Similarly, displaying a portion of a coronary artery in axial, sagittal, and coronal views in addition to CMPR, SMPR, and cross-sectional views can provide further information to an analyst that would not otherwise be perceptible using fewer views of the coronary artery. In various embodiments, any of the information described or illustrated herein as determined by the system or an analyst interacting with the system, as well as other information related to the coronary arteries / vessels associated with the set of CT images ("arterial information"), including information indicative of stenosis and plaque in segments of the coronary vessels in the set of CT images (e.g., from another external source, e.g., the analyst), as well as information indicative of the identification and location of the coronary vessels in the set of CT images, may be stored in the system and presented in various panels and reports of the user interface. The present disclosure allows for easier and faster analysis of a patient's coronary arteries and features associated with the coronary arteries. The present disclosure also allows for faster analysis of coronary artery data by allowing for quick and accurate access to selected portions of the coronary artery data. Without the use of the systems and methods of the present disclosure, the quick selection, display, and analysis of CT images and coronary artery information may be cumbersome and inefficient, and may lead to an analyst missing important information in the analysis of a patient's coronary arteries, which may lead to an inaccurate assessment of the patient's condition.

[0363] In various embodiments, the system can identify the patient's coronary arteries automatically (e.g., using a machine learning algorithm during a pre-processing step of a set of CT images associated with the patient) or interactively (e.g., by receiving at least some input from a user) by an analyst or practitioner using the system. As described herein, in some embodiments, processing the raw CT scan data can include analyzing the CT data to determine and / or identify the presence and / or absence of specific arterial vessels in the patient's body. As a naturally occurring phenomenon, a specific artery may be present in a particular patient's body while such specific artery may not be present in another patient's body. In some embodiments, the system can be configured to identify and label the arterial vessels detected in the scan data. In certain embodiments, the system can be configured to allow a user to click on the label of an identified artery in the patient's body, thereby enabling the artery to be highlighted in an electronic representation of multiple arterial vessels present in the patient's body. In some embodiments, the system can be configured to analyze the arteries present in the CT scan data and display various views of the arteries present in the patient's body, for example, in 10-15 minutes or less. In contrast, as an example, visual assessment of a CT scan to identify only stenosis, without considering benign or malignant plaque or any other factors, can take anywhere from 15 minutes to over an hour depending on skill level and can have significant variability across radiologists and / or cardiac imaging devices.

[0364] Some systems allow the analyst to view CT images associated with the patient, but do not have the ability to display all the necessary views in real time or near real time, including 3-D arterial tree views, multiple SMPR views, and cross-sectional views of the patient's coronary arteries, as well as corresponding axial, sagittal, and / or coronary views. An embodiment of the system can configure this to display one or more or all of the views, providing unparalleled visibility of the patient's coronary arteries, allowing the analyst or practitioner to perceive features and information that may not be simply perceptible without these views. That is, a user interface configured to show all of these views, as well as information related to the displayed coronary vessels, allows the analyst or practitioner to use his or her own experience in conjunction with the information provided by the system to better identify arterial conditions, which may help in making treatment decisions for the patient. In addition, the information determined by the system and displayed by the user interface that cannot be perceived by the analyst or practitioner is presented in a manner that can be easily understood and quickly incorporated. As an example, knowledge of the actual radiodensity value of the plaque is not determined by the analyst simply looking at the CT image, but the system can present that a complete analysis of all plaque has been found.

[0365] In general, arterial vessels are curvilinear in nature. Thus, the system can be configured to straighten such curvilinear arterial vessels into a substantially linear arterial view, which in some embodiments is referred to as a linear multiplanar reconstruction (MPR) view. In some embodiments, the system can be configured to present a dashboard view showing multiple arterial vessels in a linear multiplanar reconstruction view. In some embodiments, the linear view of the arterial vessel shows a cross-sectional view along the longitudinal axis of the arterial vessel (or the vessel's length or long axis). In some embodiments, the system can be configured to allow the user to rotate 360° about the longitudinal axis of the substantially linear arterial vessel to inspect the vessel wall from various perspectives and angles. In some embodiments, the system is configured to show not only the stenosis of the vessel inner diameter, but also the characteristics of the vessel inner and / or outer walls themselves. In some embodiments, the system can be configured to display multiple arterial vessels in multiple linear views, e.g., in an SMPR view.

[0366] In some embodiments, the system can be configured to show multiple arterial vessels in a perspective view to better show the curvature of the arterial vessels to the user. In some embodiments, the perspective view is referred to as a curved multiplanar reconstruction. In some embodiments, the perspective view includes a CT image of the heart and vessels, for example in the form of a wooden image of the arteries. In some embodiments, the perspective view includes a modified CT image showing the arterial vessels that does not display the heart tissue to better highlight the heart vessels. In some embodiments, the system can be configured to allow the user to rotate the perspective view to view the patient's various arteries from different perspectives. In some embodiments, the system can be configured to show a cross-sectional view of the arterial vessels along the horizontal axis (or width or short axis of the vessel). In contrast to a cross-sectional view along the vertical axis, in some embodiments, the system can allow the user to more clearly see the stenosis or narrowing of the vessel wall by viewing the arterial vessel from a cross-sectional view across the horizontal axis.

[0367] In some embodiments, the system is configured to display a plurality of arterial vessels in a cartoon, illustrated, or cartoon image. In the illustrated image of the arterial vessels, in some embodiments, the system may utilize solid or gray-scaling of a particular arterial vessel or a segment of a particular arterial vessel to indicate various degrees of risk of a cardiovascular event occurring in the particular arterial vessel or segment of an arterial vessel. For example, the system may be configured to display a first arterial vessel in yellow to indicate a medium risk of a cardiovascular event occurring in the first arterial vessel and a second arterial vessel in red to indicate a high risk of a cardiovascular event occurring in the second arterial vessel. In some embodiments, the system may be configured to allow a user to interact with various arterial vessels and / or segments of arterial vessels to better understand a specified risk associated with the arterial vessel or segment of an arterial vessel. In some embodiments, the system may allow a user to switch from the illustrated image of the patient's arteries to a CT image.

[0368] In some embodiments, the system may be configured to display all or a portion of the various views described herein in a single dashboard view. For example, the system may be configured to display a linear view along with a perspective view. In another example, the system may be configured to display a linear view along with an illustrated image.

[0369] In some embodiments, the processed CT image data may enable the system to utilize such processed data to display various arteries of a patient to a user. As described above, the system may be configured to utilize the processed CT data to generate a linear view of a patient's multiple arterial vessels. In some embodiments, the linear view displays the patient's arteries in a linear fashion that mimics a substantially straight line. In some embodiments, generating the linear view requires stretching the image of one or more naturally occurring curved arterial vessels. In some embodiments, the system may be configured to utilize such processed data to enable a user to rotate the displayed linear view of the artery in a 360° rotatable manner. In some embodiments, the processed CT image data may enable visualization and comparison of arterial morphology over time, i.e., throughout the cardiac cycle. Arterial dilation, or lack thereof, may represent a healthy artery versus a diseased artery that is unable to vasodilate. In some embodiments, a predictive algorithm may determine whether an artery is capable of dilating by simply examining a single time point.

[0370] As described above, aspects of the system can aid in visualizing a patient's coronary arteries. In some embodiments, the system can be configured to utilize processed data from raw CT scans to dynamically generate a visualization interface for a user to interact with and / or analyze a particular patient's data. The visualization system can display multiple arteries associated with the patient's heart. The system can be configured to display multiple arteries in a substantially linear manner, even though arteries are not linear within the patient's body. In some embodiments, the system can be configured to allow a user to scroll up and down or left and right along the length of the artery to visualize different extents of the artery. In some embodiments, the system can be configured to allow a user to rotate the artery 360 degrees to allow a user to view different portions of the artery at different angles.

[0371] Advantageously, the system may be configured to include or generate markings on areas where there is an amount of plaque accumulation above a threshold level. In some embodiments, the system may be configured to allow a user to target specific areas of an artery for further inspection. The system may be configured to allow a user to click on one or more marked areas of an artery to display underlying data associated with the artery at a particular point along the length of the artery. In some embodiments, the system may be configured to generate a cartoon rendition of the patient's artery. In some embodiments, the cartoon or computer-generated representation of the artery may include a color-coding scheme to highlight specific areas of the patient's artery for further inspection by the user. In some embodiments, the system may be configured to generate a cartoon or computer-generated image of the artery using the color red or any other graphic representation to represent arteries requiring further analysis by the user. In some embodiments, the system may label the cartoon representation of the artery and the 3D representation of the artery described above with stored coronary vessel labels according to a labeling scheme. If desired by the user, the labeling scheme may be modified or refined, and preferred labels may be stored and used to label the coronary arteries.

[0372] In some embodiments, the system can be configured to identify areas of an artery where ischemia is likely to be found. In some embodiments, the system can be configured to identify areas of plaque where malignant plaque is present. In some embodiments, the system can be configured to identify areas of malignant plaque by determining whether the shade and / or level of gray scale of the areas in the artery exceeds a threshold level. In one example, the system can be configured to identify areas of plaque where an image of the plaque area is black or substantially black or dark gray. In one example, the system can be configured to identify areas of "benign" plaque by designating the whiteness or light grayness of the plaque area in the artery.

[0373] In some embodiments, the system is configured to identify portions of arterial vessels at high risk for cardiac events and / or to draw a contour that follows the vessel wall or a profile of plaque accumulation along the vessel wall. In some embodiments, the system is further configured to display this information to the user and / or provide an editing tool for the user to modify the designation of the identified portions or contours if the user believes that the designation of the contour drawn by the AI ​​algorithm is inappropriate. In some embodiments, the system includes an editing tool called "snap-to-lumen", where the user selects an area of ​​interest by drawing a box around a particular area of ​​the vessel and selecting the snap-to-lumen option, and the system automatically redraws the designation of the contour to more closely follow the borders of the vessel wall and / or plaque accumulation, using image processing techniques such as, but not limited to, edge detection. In some embodiments, the AI ​​algorithm does not process the medical image data with perfect accuracy, and therefore the editing tool is necessary to complete the analysis of the medical image data. In some embodiments, the final user editing of the medical image data allows for faster processing of the medical image data than would be possible if the medical image data were processed using only the AI ​​algorithm.

[0374] In some embodiments, the system is configured to replicate images from higher resolution imaging. As an example, in CT, partial volume artifacts from calcium are known artifacts of CT, resulting in overestimation of calcium volume and narrowing of the artery. By training and evaluating CT arterial appearance against that of intravascular ultrasound or optical coherence tomography or histopathology, in some embodiments, the CT arterial appearance may be replicated to be similar to IVUS or OCT, thus de-blooming coronary calcium artifacts and improving the accuracy of the CT images.

[0375] In some embodiments, the system is configured to provide a graphical user interface that displays a vessel from a beginning segment to an end segment, and / or the tapering of the vessel along the course of the vessel length. Many examples of panels that may be displayed in the graphical user interface are illustrated and described with reference to Figures 6A-9N. In some embodiments, the portions, panels, buttons, or information displayed in the user interface are arranged differently than described herein and illustrated in the figures. For example, a user may have a preference to have different views of arteries placed in different portions of the user interface.

[0376] In some embodiments, the graphical user interface is configured to annotate the displayed vascular image with plaque accumulation data obtained from the AI ​​algorithmic analysis to indicate stenosis or stenosis images of the blood vessel. In some embodiments, the graphical user interface system is configured to annotate the displayed vascular image with color markings or other markings to indicate areas of high risk or areas for further analysis, areas of moderate risk, and / or areas of low risk. For example, the graphical user interface system may be configured to annotate certain areas along the blood vessel length with red markings or other graphic markings to indicate significant malignant fatty plaque accumulation and / or stenosis. In some embodiments, the annotated markings along the blood vessel length are based on one or more variables, such as, but not limited to, stenosis, biochemical tests, biomarker tests, AI algorithmic analysis of medical imaging data, and / or others. In some embodiments, the graphical user interface system is configured to annotate the vascular image with atherosclerosis images. In some embodiments, the graphical user interface system is configured to annotate the vascular image with ischemia images. In some embodiments, the graphical user interface is configured to allow the user to rotate the blood vessel 180 degrees or 360 degrees to view the blood vessel and annotated plaque accumulation image from different angles. From this image, the user can manually determine a stent length and diameter to address the stenosis, and in some embodiments, the system is configured to analyze the medical image information to determine a recommended stent length and diameter and display the stent proposed to be implanted in the graphical user interface to show the user how the stent will address the stenosis in the identified region of the blood vessel. In some embodiments, the systems, methods, and devices disclosed herein can be applied to other regions of the subject's body and / or other blood vessels and / or organs, whether the subject is a human or other mammal.

[0377] [Example] One of the main uses of such a system may be to determine the presence of plaque in blood vessels, such as but not limited to coronary vessels. The type of plaque may be visualized based on Hounsfield unit density to improve user readability. Embodiments of the system also provide quantification of variables related to stenosis and plaque composition at both vessel and lesion level for segmented coronary arteries.

[0378] In some embodiments, the system is configured as a web-based software application intended for use by trained medical professionals as an interactive tool to review and analyze cardiac CT data to determine the presence and extent of coronary plaque (i.e., atherosclerosis) and stenosis in patients undergoing coronary computed tomography angiography (CCTA) to evaluate for coronary artery disease (CAD) or suspected CAD. The system post-processes CT images obtained using a CT scanner. The system is configured to generate a user interface that provides tools and functionality for characterization, measurement, and visualization of coronary artery features.

[0379] Features of embodiment of the system may include, for example, centerline and lumen / vessel extraction, plaque composition overlay, user identification of stenosis, real-time calculated vessel statistics (including vessel length, lesion length, vessel volume, lumen volume, plaque volume (non-calcified, calcified, low-density non-calcified plaque, and total)), maximum remodeling index, and area / diameter stenosis ratio (e.g., percentage), two-dimensional (2D) visualization of multiplanar reformatted vessels and cross-sectional images, interactive three-dimensional (3D) rendered coronary artery trees, animated arterial tree visualization corresponding to actual vessels appearing in CT images, user-modifiable semi-automatic vessel segmentation, and user identification of stents and chronic total occlusions (CTOs).

[0380] In one embodiment, the system uses 18 coronary segments in the coronary vessel tree (e.g., following Society of Cardiovascular Computed Tomography guidelines). The coronary segment labels include: ·pRCA-proximal right coronary artery mRCA - Middle right coronary artery dRCA - Distal right coronary artery ·R-PDA-Right posterior descending artery ·LM-Left aorta ·pLAD-proximal left anterior descending artery ·mLAD-middle left anterior descending artery · dLAD-Distal left anterior descending artery D1 - 1st diagonal D2 - 2nd diagonal pCx - proximal left circumflex artery OM1 - First blunt edge ·LCx-distal left rotation ·OM2-2nd obtuse edge ·L-PDA-Left posterior descending artery R-PLB-Right posterior lateral branch ·RI-Intermediate branch artery ·L-PLB-Left posterior wall branch

[0381] Other embodiments may include more or fewer coronary artery segment labels. The coronary artery segments present in an individual patient depend on whether the patient has right or left coronary artery dominance. Some segments are present only when the patient has right coronary artery dominance, and some segments are present only when the patient has left coronary artery dominance. Thus, in many cases, if not all, a patient will not have all 18 segments. The system takes into account most known variations.

[0382] In one implementation of the system, a CT scan was processed by the system and the resulting data was compared with ground truth results generated by expert readers. The Pearson correlation coefficients and Bland-Altman agreement between the system results and the expert reader results are shown in the table below. [Table 1]

[0383] 6A-9N show one embodiment of the system's user interface, showing example panels, graphics, tools, representations of CT images, and properties, structures, and statistics associated with coronary vessels found in a set of CT images. In various embodiments, the user interface is flexible and can be configured to show various configurations of panels, images, graphic representations of CT images, and properties, structures, and statistics; for example, based on the analyst's preferences. The system has multiple menus and navigation tools to aid in visualization of the coronary arteries. Keyboard and mouse shortcuts can also be used to navigate images and information associated with a patient's set of CT images.

[0384] FIG. 6A illustrates an example of a user interface 600 having multiple panels (views) that can be generated and displayed on a CT image analysis system described herein and can show various corresponding views of a patient's arteries and information about the arteries. In one embodiment, the user interface 600 illustrated in FIG. 6A can be a starting point for an analysis of a patient's coronary arteries and may be referred to herein as a "study page" (or study page 600). In some embodiments, the study page can include a number of panels that can be positioned in different locations of the user interface 600, for example, based on an analyst's preferences. In various examples of the user interface 600, a particular panel can be selected for display (e.g., based on user input) from among the possible panels that may be displayed.

[0385] The example study page 600 shown in FIG. 6A includes an arterial tree 602 including a three-dimensional (3D) representation of the coronary vessels based on a CT image, and includes a first panel 601 (also indicated with a circled "2") depicting the coronary vessels identified in the CT image and further depicting their respective segment labels. While processing the CT image, the system can determine the extent of the coronary vessels and an arterial tree is generated. Structures that are not part of the coronary vessels (e.g., cardiac tissue and other tissues around the coronary vessels) are not included in the arterial tree 602. Thus, the arterial tree 602 in FIG. 6A does not include cardiac tissue between the branches (vessels) 603 of the arterial tree 602, allowing all parts of the arterial tree 602 to be visualized without being obstructed by cardiac tissue.

[0386] The example study page 600 also includes a second panel 604 (also indicated with a circled "1a") showing at least a portion of a selected coronary vessel in at least one linear multiplanar reconstruction (SMPR) vessel image. The SMPR image is an elevation view of the vessel at a particular rotational aspect. Although multiple SMPR images are displayed in the second panel 604, each image may be at a different rotational aspect. For example, at any 1°, or at 0.5°, from 0° to 259.5°, 360° is the same view as 0°. In this example, the second panel 604 includes four linear multiplanar vessels 604a-d, displayed in elevation at relative rotations of 0°, 22.5°, 45°, and 67.5°, where the rotations show the upper portion of the linear multiplanar vessel. In some embodiments, the rotation of each image may be selected by the user, for example, at different relative rotation intervals. The user interface illustrates a rotation tool 605 configured to receive input from a user and can be used to adjust the rotation of the SMPR image (e.g., by one or more degrees). One or more graphics related to the vessel shown in the SMPR image can also be displayed, such as a graphic representing the lumen of the vessel, a graphic representing the vessel wall, and / or a graphic representing plaque.

[0387] This example study page 600 also includes a third panel 606 (also indicated by a circled "1c") configured to show a cross-sectional view of a blood vessel 606a generated based on a CT image in the set of CT images for the patient. The cross-sectional view corresponds to the blood vessel shown in the SMPR image. The cross-sectional view also corresponds to a location indicated by a user (e.g., with a pointing device) in the blood vessel in the SMPR image. The user interface is configured such that selection of a particular location along the coronary vessel in the second panel 604 causes an associated CT image to be displayed in the cross-sectional view in the third panel 606. In this example, a graphic 607 is displayed in the second panel 604 and the third panel 606 shows the extent of plaque in the blood vessel.

[0388] This example study page 600 also includes a fourth panel 608 that includes anatomical planar views of a selected coronary vessel. In this embodiment, the study page 600 includes an axial view 608a (also indicated by a circled "3a"), a coronal view 608b (also indicated by a circled "3b"), and a sagittal view 608c (also indicated by a circled "3c"). The axial view is a cross-sectional or "top" view. The coronal view is a frontal view. The sagittal view is a lateral view. The user interface is configured to display a corresponding view of the selected coronary vessel, e.g., a view of the selected coronary vessel at a location on the coronary vessel selected by the user (e.g., one of the SMPR views in the second panel 604).

[0389] FIG. 6B shows another example of a study page (user interface) 600 having multiple panels that can be generated and displayed on the system and can show various corresponding views of the patient's arteries. In this example, the user interface 600 displays a 3D arterial tree in a first panel 601, a cross-sectional view in a third panel 606, and axial, coronal, and sagittal views in a fourth panel 608. Instead of the second panel 604 shown in FIG. 6A, the user interface 600 includes a fifth panel 609 that shows a curved multiplanar reconstruction (CMPR) vessel view of a selected coronary vessel. The fifth panel 609 can be configured to show one or more CMPR views. In this example, two CMPR views are generated, with a first CMPR view 609a displayed at 0° and a second CMPR view 609b displayed at 90°. The CMPR views can be generated and displayed at various relative rotations, for example, from 0° to 259.5°. The coronary vessels shown in the CMPR image correspond to the selected vessel and to the vessels displayed in the other panels. When a location on a vessel is selected in one panel (e.g., the CMPR image), the images in the other panels (e.g., cross-sectional, axial, sagittal, and coronal images) can be automatically updated to also show the vessel at that selected location in their respective images, thereby greatly improving the information presented to the user and increasing the efficiency of the analysis.

[0390] 6C, 6D, and 6E show certain details of the multiplanar reconstruction (MPR) vessel image in the second panel, and certain functionality associated with this image. After the user verifies the accuracy of the segmentation of the coronary artery tree in panel 602, he can proceed to interact with the MPR image, where he can edit the individual vessel segments (vessel wall, lumen, etc.). In the SMPR and CMPR images, the vessel can be rotated incrementally (e.g., 22.5°) by using the arrow icons 605 shown in Figs. 6C and 6D. Alternatively, the vessel can be rotated 360° in 1 degree increments by using the rotate command 610 as shown in Fig. 6E. The vessel can also be rotated in the user interface 600 by pressing the COMMAND or CTRL button and left-clicking+dragging the mouse.

[0391] FIG. 6F shows further information of a three-dimensional (3D) rendering of the coronary artery tree 602 in a first panel 601, which allows the user to view the vessels and modify the vessel labels. FIG. 6G shows shortcut commands for the coronary artery tree 602, axial view 608a, sagittal view 608b, and coronal view 608c. In the panel 601 shown in FIG. 6F, the user can rotate the arterial tree and zoom in and out of the 3D rendering using commands selected in the user interface shown in FIG. 6G. By clicking on a vessel, it turns yellow, indicating that it is the vessel currently being inspected. In this view, the user can rename or erase the vessel by right-clicking on the vessel name to open a panel 611 configured to receive input from the user to rename the vessel. The panel 601 also includes a control that can be activated to turn the displayed labels "on" or "off." Figure 6H further shows a user interface panel 608 for viewing DICOM images in three anatomical planes: axial, coronal, and sagittal. Figure 6I shows panel 606 showing cross-sectional images of blood vessels. Scrolling, zooming in / out, and panning commands can also be used on these images.

[0392] 6J and 6K illustrate certain aspects of functionality in the toolbar 612 and menu navigation of the user interface 600. FIG. 6J illustrates a toolbar of the user interface for navigating vessels. The toolbar 612 includes buttons 612a, 612b, etc. for each vessel displayed on the screen. The user interface 600 is configured to display buttons 612a-n to indicate various information to the user. In one example, when a vessel is selected, the corresponding button, such as button 612c, is highlighted (e.g., displayed in yellow). In another example, a dark gray button with white text indicates that the vessel is available for analysis. In one example, a shaded black button 612d means that the vessel could not be analyzed by the software because it is not anatomically present or has too many artifacts. A gray button 612e with a check mark indicates that the vessel has been examined.

[0393] FIG. 6K shows a diagram of the user interface 600 including an expanded menu for viewing all the series (of images) available for review and analysis. If the system provides more than one of the same vessel segments for analysis from different series of images, the user interface is configured to receive user input to select the desired series for analysis. In one example, input can be received indicating the series to review by selecting one of the radio buttons 613 from the series of interest. The radio button changes from gray to purple when selected for review. In one embodiment, the software selects the two series of highest diagnostic quality for analysis by default, but all series are available for review. The user can use clinical judgment to determine if the series selected by the system are of the diagnostic quality required for analysis, and should select different series for analysis if desired. The series selected by the system shall improve workflow by prioritizing diagnostic quality images. The system shall not interchange the user's review of all series and selection of diagnostic quality images within a study. The user can submit any sequence shown in FIG. 6K to the system and suggest vessel segmentation by hovering the mouse over the sequence and selecting the “Analyze” button 614 as shown in FIG. 6L.

[0394] FIG. 6M illustrates a panel that may be displayed on the user interface 600 for adding a new vessel on an image, according to one embodiment. To add a new vessel on an image, the user interface 600 may receive user input via a “+Add Vessel” button on the toolbar 612. The user interface displays a “Create Mode” 615 button that appears on a fourth panel 608 on the axial, coronal, and sagittal views. A vessel may then be added on the image by scrolling and clicking the left mouse button to create multiple dots (e.g., green dots). Once the new vessel has been added, it is previewed as a new vessel in the MPR, cross-section, and 3D arterial tree views. The user interface is configured to receive a “Done” command indicating that the vessel addition is complete. Next, to segment the vessel utilizing the semi-automatic segmentation tool of the system, “Analyze” is clicked on the toolbar, and the user interface displays the proposed segmentation for inspection and correction. A vessel name can be chosen by selecting "New" on the 3D artery tree in the first panel 601, thereby activating a name panel 611, a vessel name can be selected from the panel 611, and then the new vessel and its name are stored. In one embodiment, if the software cannot identify the vessel being added by the user, it reverts to a straight vessel line connecting the green dots added by the user, and the user can adjust the center line. The user interface pop-up menu 611 allows for the identification and naming of new vessels quickly and consistently according to a standard format.

[0395] FIG. 7A shows an example of an editing toolbar 714 that includes editing tools that allow a user to process a CT scan with a machine learning algorithm and then allow an analyst to modify and improve the accuracy of findings resulting from processing the CT scan and the information generated by the machine learning algorithm. In some embodiments, the user interface includes editing tools that can be used to modify and improve the accuracy of findings. In some embodiments, the editing tools are located on the left side of the user interface, as shown in FIG. 7A. Below is a list and description of the available editing tools. Hover over each button (icon) to display the name of each tool. These tools can be activated and deactivated by clicking on them. If the tool is gray in color, it is deactivated. If the software has identified any of these characteristics in the vessel, an annotation will already be on the image when the tool is activated. The editing tools in the toolbar can include one or more of the following tools: Lumen Wall 701, Snap-To Vessel Wall 702, Vessel Wall 703, Snap-To Lumen Wall 704, Segment 705, Stenosis 706, Plaque Overlay 707, Centerline 708, Chronic Total Occlusion (CTO) 709, Stent 710, Exclusion 711, Tracker 712, and Distance 713. User interface 600 is configured to activate each of these tools by receiving a user selection of a respective tool icon (shown in the table below and in FIG. 7A ) and is configured to provide the functionality described in the Edit Tool Description Table below. [Table 2]

[0396] 7B and 7C illustrate the specific functionality of the tracker tool. The tracker tool 712 allows the user to orient and correlate images shown in the various panels of the user interface 600, e.g., SMPR, CMPR, cross-sectional, axial, coronal, sagittal, and 3D arterial tree views. To activate, the tracker icon is selected on the editing toolbar. When the tracker tool 712 is activated, the user interface generates and displays a line 616 (e.g., a red line) on the SMPR or CMPR view. The system generates a corresponding (red) disk 617 on the user interface, which is displayed in the corresponding position to the line 616 on the 3D arterial tree of the first panel 601. The system generates a corresponding (red) dot on the user interface, which is displayed in the corresponding position to the line 616 on the axial, sagittal, and coronal views of the fourth panel. Line 616, disk 617, and dot 618 are all position indicators that refer to the same location in different views, such that scrolling up or down on any of the trackers results in the same movement of the position indicators in the other views. User interface 600 also displays in panel 606 cross-sectional images that correspond to the locations indicated by the position indicators.

[0397] 7D and 7E illustrate the specific functionality of the vessel and lumen wall tools used to modify lumen and vessel wall contours. The lumen wall tool 701 and vessel wall tool 703 are configured to modify lumen and vessel wall (also referred to herein as contours, boundaries, or features) previously determined for a vessel (e.g., determined by processing a CT image using a machine learning process). These tools are used by the system to determine measurements that are output or displayed. By interacting with the contours generated by the system with these tools, the user can improve the accuracy of the location of the contours, and any measurements derived from those contours. These tools can be used in SMPR and cross-sectional images. The tools are activated by selecting the vessel and lumen icons 701, 703 on the editing toolbar. The vessel wall 619 is displayed in a graphical "trace" overlay in one color (e.g., yellow) in the MPR and cross-sectional images. The lumen wall 629 is displayed in a graphical "trace" overlay in a different color (e.g., purple). In one embodiment, the user interface is configured to refine the contours through user interaction. For example, to refine a contour, a user can hover a pointing device (e.g., mouse, stylus, finger) over the contour to highlight it, click on the desired vessel or lumen wall contour, and drag the displayed tracing to a different location to set a new boundary. The user interface 600 is configured to automatically save any changes to these tracings. The system recalculates any measurements derived from the contour changes in real-time or near real-time. Also, changes made in one panel of one image are correspondingly displayed in the other image / panel.

[0398] FIG. 7F shows the lumen wall button 701 and the snap-to vascular wall button 702 (left) and the vascular wall button 703 and the snap-to luminal wall button 704 (right) of the user interface 600, which can be used to activate the lumen wall / snap-to vascular tools 701, 702 and the vascular wall / snap-to luminal wall tools 703, 704, respectively. The user interface provides these tools to modify previously determined lumen and vascular wall contours. The snap-to vascular / lumen wall tools are used to easily and quickly close gaps between the lumen and vascular wall contours, i.e., move the lumen contour trace and the vascular contour trace to be the same or substantially the same, reducing interactive editing time. The user interface 600 is configured to activate these tools when the user hovers a pointing device over the tools, thereby causing the snap-to button to appear. For example, when hovering over the lumen wall button 701, the snap-to-vessel button 702 appears to the right of the lumen wall button, and when hovering over the vessel wall button 703, the snap-to-lumen wall button 704 appears next to the vessel wall button 703. A button is selected to activate the desired tool. Referring to FIG. 7G, the pointing device can be used to click and drag along the intended portion of the vessel to edit to a second point 621, and a range 622 appears indicating where the tool will operate. Once the end of the desired range 622 has been drawn, the lumen and vessel wall will stay together by releasing the selection.

[0399] FIG. 7H shows an example of the second panel 602 that can be displayed while using the segment tool 705, which allows for marking the boundaries between individual coronary artery segments on the MPR. The user interface 600 is configured such that when the segment tool 705 is selected, a line (e.g., lines 623, 624) appears on the vessels in the SMPR image and on the vessel image of the second panel 602. The lines indicate the segment boundaries determined by the system. Names are displayed with icons 625, 626 adjacent to the respective lines 623, 624. To edit the names of the segments, click on the icons 625, 626 and label accordingly using the name panel 611 shown in FIG. 7I. Segments can also be erased, for example, by selecting the trash icon. The lines 623, 624 can be moved up or down to define the segment of interest. If there are no segments, the user can add a new segment using the add segment button and label it using the labeling function of the segment labeling pop-up menu 611.

[0400] 7J-7M show an example of using the stenosis tool 706 on the user interface 600. For example, FIG. 7L shows a stenosis button that can be used to drop stenosis markers based on the user edited lumen and vessel wall contours. FIG. 7M shows stenosis markers on a segment in a curved multiplanar vessel (CMPR) image. A second panel 604 can be displayed while using the stenosis tool 706 that allows the user to indicate markers that mark the extent of stenosis in the vessel. In one embodiment, the stenosis tool includes a set of five markers that are used to mark the extent of stenosis in the vessel. These markers are defined as follows: R1: Proximal normal slice closest to stenosis / lesion P: Abnormal slice most proximal to stenosis / lesion O: slice with maximum occlusion D: Abnormal slice most distal to stenosis / lesion R2: Distal normal slice closest to stenosis / lesion

[0401] In one embodiment, there are two ways to add stenosis markers to multiplanar images (straight and curved). After selecting the stenosis tool 706, a stenosis can be added by activating the stenosis button shown in FIG. 7K or FIG. 7L; to drop five evenly spaced stenosis markers, (i) click the stenosis "+" button (FIG. 7K); (ii) a series of five evenly spaced yellow lines will appear on the vessel and the user must edit these markers to their applicable positions; (iii) move all five markers simultaneously by clicking inside the highlighted area enclosed by the markers and dragging them up or down; (iv) move individual markers by clicking individual yellow lines or tags and moving them up or down; and (v) click the red trash icon to erase the stenosis. To drop stenosis markers based on the user edited lumen and vessel wall contours, click the stenosis "+" button (FIG. 7K).

number

[0402] FIG. 7N shows an example of a panel that may be displayed while using the plaque overlay tool 707 of the user interface. In one embodiment, referring to FIG. 7N, "plaque" is categorized as low density non-calcified plaque (LD-NCP) 701, non-calcified plaque (NCP) 632, or calcified plaque (CP) 633. The tool is activated by selecting the plaque overlay tool 707 in the edit toolbar. When activated, the plaque overlay tool 707 overlays different colors on the vessel in the SMPR image in the second panel 604, the cross-section of the SMPR, and the cross-section image in the third panel 606 (see, e.g., FIG. 7R) with the range of plaque based on Hounsfield Units (HU) density. In addition, a legend opens in the cross-section image to correspond the plaque type to the plaque overlay color as shown in FIG. 7O and FIG. 7Q. The user can select different HU ranges for the three different types of plaque by clicking the "Edit Thresholds" button located in the upper right corner of the cross-sectional image as shown in Figure 7P. In one embodiment, the default plaque threshold settings for the values ​​are shown in the table below. [Table 3]

[0403] The default value can be corrected, if desired, using, for example, the plaque threshold interface shown in FIG. 7Q. Although a default value is provided, the user can select a different plaque threshold value based on their clinical judgment. The user can further inspect the area of ​​interest using the cross-sectional image of the third panel 606 shown in FIG. 7R. The user can also view the selected plaque threshold value in the vascular statistics panel of the user interface 600 shown in FIG. 7S.

[0404] The centerline tool 708 allows the user to adjust the center of the lumen. Changing the center point (of the centerline) may change the quantification of the lumen and vessel wall, as well as plaque, if present. The centerline tool 708 is activated by selecting it on the user interface 600. A line 635 (e.g., a yellow line) appears in the CMPR image 609 and a point 634 (e.g., a yellow point) appears in the cross-sectional image in the third panel 606. The centerline can be adjusted as needed by clicking and dragging the line / point. Any changes in the CMPR image are reflected in the cross-sectional image and vice versa. The user interface 600 provides several ways to extend the centerline of an existing vessel. For example, the user can extend the centerline as follows: (1) right-click on the outlined vessel dot 634 on the axial, coronal, or sagittal view (see Figure 7U); (2) select "Extend from Beginning" or "Extend from End" (see Figure 7U) so that the view jumps to the beginning or end of the vessel; (3) extend the vessel by adding a (green) dot (see Figure 7V); and (4) when finished, select the (blue) check mark button to stop the extension and select the (red) "x" button (see, e.g., Figure 7V). The user interface then extends the vessel according to the changes made by the user. The user then manually edits the lumen and vessel walls on the SMPR or cross-sectional view (see, e.g., Figure 7W). If the user interface cannot identify the vessel segment being added by the user, it reverts to a straight vessel line connecting the dots added by the user. The user can then adjust the centerline.

[0405] The user interface 600 also provides a chronic total occlusion (CTO) tool 709 that identifies portions of an artery with a CTO, i.e., a portion of the artery that has 100% stenosis and no detectable blood flow. Plaque within a CTO is not included in the overall plaque quantification because it is more likely to contain a large amount of thrombus. To activate, click on the CTO tool 709 in the editing toolbar 612. To add a CTO, click on the CTO "+" button in the user interface. Two lines (markers) 636, 637 appear in the MPR image in the second panel 604, as shown in FIG. 7X, which shows a portion of the vessel with a CTO. The markers 636, 637 can be moved to adjust the extent of the CTO. If more than one CTO is present, additional CTOs can be added by again activating the CTO "+" button in the user interface. CTOs can also be erased if necessary. The location of the CTO is stored. In addition, the portion of the vessel that is within a specified CTO is not included in the overall plaque calculation, and plaque quantification determinations are recalculated as necessary after a CTO is identified.

[0406] The user interface 600 also provides a stent tool 710 that indicates where a stent is present within the vessel. The stent tool is activated by user selection of the stent tool 710 in the toolbar 612. To add a stent, click on the stent "+" button provided in the user interface. Two lines 638, 639 (e.g., purple lines) appear in the MPR image as shown in FIG. 7Y, and the lines 638, 639 can be moved to indicate the extent of the stent by clicking on the individual lines 638, 639 and moving them up or down along the vessel to the end of the stent. Overlap with stent (or CTO / exclusion / stenosis) markers is not allowed by the user interface 600. Stents can also be erased.

[0407] The user interface 600 also provides an exclusion tool 711 configured to indicate portions of the vessel to be excluded from analysis due to blurring caused by motion, contrast, misalignment, or other reasons. By excluding low quality images, the overall quality of the analysis results for the non-excluded portions of the vessel is improved. To exclude top or bottom portions of the vessel, the segment tool 705 and the exclusion tool 711 of the editing toolbar 612 are activated. FIG. 7Z illustrates the use of the exclusion tool to exclude a portion from the top of the vessel. FIG. 7AA illustrates the use of the exclusion tool to exclude a bottom portion of the vessel. The first segment marker acts as an exclusion marker for the top portion of the vessel. The area enclosed by the exclusion marker is excluded from all statistical calculations of the vessel. Areas can be excluded by dragging the top segment marker to the bottom of the desired excluded area. The excluded area is highlighted. Alternatively, the "end" marker can be dragged to the top of the desired excluded area. The excluded area is highlighted and the user can enter a reason for the exclusion into the user interface (see FIG. 7AC). To add a new exclusion to the center of the vessel, activate the exclusion tool 711 in the editing toolbar 612. Click on the exclusion "+" button. A user interface pop-up window appears (FIG. 7AC) regarding the reason for the exclusion, where the reason can be entered and stored with reference to the exclusion area shown. Two markers 640, 641 appear in the MPR as shown in FIG. 7AB. Move both markers simultaneously by clicking inside the highlighted area. The user can move individual markers by clicking and dragging the lines 640, 641. The user interface 600 tracks the location of the exclusion marker lines 640, 641 (and any previously defined features) and prohibits overlap of the area defined by the exclusion lines 640, 641 with any previously shown portions of the vessel with a CTO, stent, or stenosis. The user interface 600 is also configured to erase a specified exclusion.

[0408] 7AD-7AG, the user interface 600 also provides a distance tool 713, which is used to measure the distance between two points on an image. This is a drag-and-drop ruler that captures precise measurements. The distance tool works in MPR, cross-sectional, axial, coronal, and sagittal views. To activate, click on the distance tool 713 in the editing toolbar 612. Then click and drag between the two desired points. A line 642 and a measurement value 643 will appear on the image displayed on the user interface 600. Clear the measurement value by right-clicking on the distance line 642 or measurement value 643 and selecting the "Delete Distance" button 644 on the user interface 600 (see FIG. 7AF). FIG. 7AD shows an example of measuring the distance of a straight multiplanar vessel (SMPR). FIG. 7AE shows an example of measuring the distance 642 of a curved multiplanar vessel (CMPR). FIG. 7AF shows an example of measuring the distance 642 of a cross-section of a vessel. FIG. 7AG shows an example of measuring distance 642 in an axial image of a patient's anatomy.

[0409] An example of a vascular statistics panel of the user interface 600 is described with reference to Figures 7AH-7AK. Figure 7AH shows the "Vascular Statistics" portion 645 of the user interface 600 (e.g., a button) of the panel that can be selected to display the vascular statistics panel 646 (or "tab"), shown in Figure 7AI. Figure 7AJ shows the specific functionality in the Vascular Statistics tab that allows the user to click through the details of multiple lesions. Figure 7AK further shows the Vascular panel that the user can use to toggle between vessels. For example, the user can hide the panel by clicking the "X" on the top right side of the panel, shown in Figure 7AI. Statistics are shown at a per vessel and per lesion (if present) level, as shown in Figure 7AJ.

[0410] If more than one lesion has been marked by the user, the user can click through the details of each lesion. To see the statistics for each vessel, the user can toggle between the vessels in the vessel panel shown in Figure 7AK.

[0411] General information regarding length and volume is presented for vessels and lesions (if present), along with plaque and stenosis information at a vessel and lesion level in the vessel statistics panel 646. The user may use the exclusion tools to exclude artifacts from the image that they do not wish to consider in the calculations. The following table shows the specific statistics available for vessels, lesions, plaque, and stenosis. [Table 4] [Table 5] [Table 6] [Table 7]

[0412] The quantitative variables used in the system and displayed in various parts of the user interface 600, for example with reference to low density non-calcified plaque, non-calcified plaque, and calcified plaque, are in Hounsfield units (HU). As is known, the Hounsfield unit scale is a quantitative scale that describes radiation and is often used with reference to CT scans as a technique for characterizing radiation attenuation, making it easier to define what a given finding represents. Hounsfield unit measurements are presented with reference to the quantitative scale. Examples of Hounsfield unit measurements for certain materials are shown in the table below. [Table 8]

[0413] In one embodiment, the information the system determines regarding stenosis, atherosclerosis, and CAD-RADS details is included on panel 800 of user interface 600, as shown in FIG. 8A. By default, the CAD-RADS score may be unselected, requiring the user to manually select a score on the CAD-RADS page. By hovering over the "#" icon, user interface 600 provides more information regarding the selected output. To see more details regarding stenosis, atherosclerosis, and CAD-RADS outputs, click on the "View Details" button at the top right of panel 800, which navigates to the available details page. In one embodiment, at the center of the centerpiece page view of user interface 600 is a non-patient specific rendition of a coronary artery tree 805 ("cartoon artery tree" 805), divided into segments 805a-805r based on SCCT coronary artery segmentation, as shown in panel 802 of FIG. 8C. All analyzed vessels are displayed in color according to legend 806 based on the maximum diameter stenosis in that vessel. Greyed out segments / vessels in cartoon arterial tree 805, e.g., segments 805q and 805r, were either anatomically unavailable or were not analyzed by the system (not all segments may be present in all patients). As shown in Figures 8B and 8C, region-wise and segment-wise information can be seen by clicking on the upper regions of the tree (RCA, LM+LAD, etc.) using, e.g., the user interface 600 selection buttons in panel 801. Alternatively, segments 805a-805r in cartoon coronary arterial tree 805 may be selected.

[0414] The stenosis atherosclerosis data displayed in the user interface in panel 807 updates accordingly as various segments are selected, as shown in FIG. 8D. FIG. 8E shows an example of a portion of the summary panel 807 by region of the user interface. FIG. 8F also shows an example of a portion of the panel 807 showing the SMPR and associated statistics of a selected vessel along the vessel at the indicated location (e.g., at the location indicated by the pointing device when moved along the SMPR visualization). That is, the user interface 600 is configured to provide plaque details and stenosis details in the SMPR visualization in panel 809 and in a pop-up panel 810 that displays information when the user interface receives location information from the user along the displayed vessel, e.g., via a pointing device. The presence of a chronic total occlusion (CT) and / or a stent is indicated at the vessel segment level. For example, FIG. 8G shows the presence of a stent in the D1 segment. FIG. 8H shows the presence of a CTO in the mRCA segment. The coronary dominance and any anomalies may be displayed below the coronary tree as shown in FIG. 8I. Anomalies selected in the analysis may be displayed, for example, by hovering a pointing device over a “details” button. If the plaque threshold was changed in the analysis, an alert may be displayed in the user interface or in the generated report indicating that the plaque threshold was changed. If anomalies are present, a coronary vessel segment 805 associated with each anomaly appears separated from the aorta as shown in FIG. 8J. In one embodiment, a textual summary of the analysis may also be displayed below the coronary tree as shown, for example, in panel 811 of FIG. 8K.

[0415] FIG. 9A shows an atherosclerosis panel 900 that may be displayed on a user interface that displays a summary of atherosclerosis information based on the analysis. FIG. 9B shows a vessel selection panel that may be used to select a vessel so that a summary of atherosclerosis information is displayed on a per-segment basis. The top section of the atherosclerosis panel 900 contains per-patient data, as shown in FIG. 9A. When the user hovers over “Segments with calcified plaque” in panel 901 or “Segments with non-calcified plaque” in panel 902, the segment names with applicable plaque are displayed. Below the patient-specific data, the user may access per-vessel and per-segment atherosclerosis data by clicking on one of the vessel buttons, as shown in FIG. 9B.

[0416] Figure 9C shows a panel 903 that can be generated and displayed on a user interface showing atherosclerosis information determined by the system on a segment-by-segment basis. The presence of positive remodeling, the highest remodeling index, and the presence of low density non-calcified plaque are reported for each segment in panel 903 shown in Figure 9C. For example, plaque data can be displayed below on a segment-by-segment basis and plaque composition volume can be displayed per segment in panel 903 shown in Figure 9C.

[0417] FIG 9D shows a panel 904 that can be displayed on the user interface that contains patient-specific data for stenosis. The top section of the stenosis panel 904 contains the patient-specific data. Further details about each count can be displayed by hovering the pointing device over the numbers, as shown in FIG 9E. The vessels included in each region are shown in the table below. [Table 9]

[0418] In one embodiment, a percentage diameter stenosis bar graph 906 can be generated and displayed in a panel 905 of the user interface, as shown in FIG. 9F. The percentage diameter stenosis bar graph 906 displays the maximum diameter stenosis of each segment. If a CTO is marked on the segment, it displays as 100% diameter stenosis. If two or more stenoses are marked on the segment, the highest value output is displayed by default, and the user can click on each stenosis bar to see the details of the stenosis and investigate smaller stenoses (if any) in that segment. The user can also scroll through each cross section by dragging the gray button in the center of the SMPR image of the vessel, as shown in FIG. 9G, to view the lumen diameter and % diameter stenosis of each cross section at any selected position.

[0419] FIG. 9H shows a panel showing one or more stenosis categories marked on the SMPR based on the analysis. Color can be used to enhance the display information. In one example, LM stenoses with ≥ 50% diameter stenosis are marked in red. For the maximum percentage diameter stenosis of each segment, as shown in panel 907 of the user interface in FIG. 9I, the reference minimum lumen diameter and lumen diameter can be displayed by "hovering" the pointing device over the graphical vessel cross-sectional representation, as shown in FIG. 9J. If a segment was not analyzed or is not anatomically present, the segment is grayed out and displays "Not Analyzed." If a segment was analyzed but no stenosis was marked, the value displays "N / A."

[0420] FIG. 9K shows a panel 908 of a user interface showing CADS-RADS score selection. The CAD-RADS panel displays the definition of CAD-RADS as defined by "Coronary Artery Disease-Reporting and Data System (CAD-RADS), an Expert Consensus Document of SCCT, ACR, and NASCI: Endorsed by ACC." The user has full control over the selection of CAD-RADS scores. In one embodiment, no score is suggested by the system. In another embodiment, a CAD-RADS score can be suggested. Once a CAD-RADS score is selected on this page, the score is displayed both in a specific user interface panel and in the full text report page. Once a CAD-RADS score is selected, the user has the option to select modifiers and a representation of the condition. Once a representation is selected, an interpretation, further cardiac testing, and management guidelines can be displayed to the user on the user interface, for example, as shown in panel 909 shown in FIG. 9L. These guidelines replicate those found in the Coronary Artery Disease-Reporting and Data System (CAD-RADS), an expert consensus document of the SCCT, ACR, and NASCI; endorsed by the ACC.

[0421] 9M and 9N show tables that can be generated and displayed in a panel of the user interface and / or included in a report. FIG. 9M shows the quantitative stenosis and vascular output. FIG. 9N shows the quantitative plaque output. These quantitative tables allow the user to inspect the quantitative per-segment stenosis and atherosclerosis output from the system analysis. The quantitative stenosis and vascular output table (FIG. 9M) includes information about the arteries and segments that were evaluated. A total is given for each vascular region. The information can include, for example, length, vessel volume, lumen volume, total plaque volume, maximum diameter percent stenosis, maximum area stenosis, and highest remodeling index. The quantitative plaque output table (FIG. 9N) includes information about the arteries and segments that were evaluated. The information can include, for example, total plaque volume, total calcified plaque volume, non-calcified plaque volume, low-density non-calcified plaque volume, and total non-calcified plaque volume. The user can also download a PDF or CSV file of the quantitative output with the full text report. The full text report presents a textual summary of atherosclerosis, stenosis, and CAD-RADS criteria. The user can edit the report as desired. If the user chooses to edit the report, the report will not automatically update the CAD-RADS selections.

[0422] 10 is a flow chart illustrating a process 1000 for analyzing and displaying CT images and corresponding information. At block 1005, the process 1000 stores computer executable instructions, a set of CT images of the patient's coronary vessels, vessel labels, and arterial information associated with the set of CT images including information of stenosis, plaque, and location of segments of the coronary vessels. All steps of the process may be performed by system embodiments described herein, for example, in the system embodiment described in FIG. 13. For example, computer executable instructions stored on one or more non-transitory computer storage media are executed by one or more computer hardware processors in communication with the one or more non-transitory computer storage media. In various embodiments, the user interface may include one or more portions or panels configured to display one or more of the images, in various views (e.g., SMPR, CMPR, cross-sectional, axial, sagittal, coronary, etc.) related to the CT images of the patient's coronary arteries, a graphical representation of the coronary arteries, features (e.g., vessel wall, lumen, centerline, stenosis, plaque, etc.) that have been extracted or corrected by a machine learning algorithm or an analyst, and information related to the CT images that has been determined by the system, by the analyst, or by the analyst's interaction with the system (e.g., measurement of features in the CT images). In various embodiments, the panels of the user interface may be arranged differently than described herein and illustrated in the corresponding drawings. A user may input to the user interface on a touch screen using a pointing device or the user's finger. In one embodiment, the user interface may receive input by determining a selection of a button / icon / portion of the user interface. In one embodiment, the user interface may receive input in a defined field of the user interface.

[0423] At block 1010, the process 1000 generates and may display in a user interface a first panel including an arterial tree 602 that includes a three-dimensional (3D) representation of the coronary vessels based on the CT image, depicting the coronary vessels identified in the CT image, and depicting segment labels, and that does not include cardiac tissue between the branches of the arterial tree. An example of such an arterial tree 602 is shown in panel 601 of FIG. 6A. In various embodiments, panel 601 may be positioned in a location of the user interface 600 other than that shown in FIG. 6A.

[0424] At block 1015, the process 1000 may receive a first input indicating a selection of a coronary vessel in the arterial tree of the first panel. For example, the first input may be received by the user interface 600 of a vessel in the arterial tree 602 of panel 601. At block 1020, in response to the first input, the process 1000 may generate and display in the user interface a second panel showing at least a portion of the selected coronary vessel in at least one straight multiplanar vascular (SMPR) image. In one example, the SMPR image is displayed in panel 604 of FIG. 6A.

[0425] At block 1025, the process 1000 may generate and display in the user interface a third panel showing a cross-sectional image of the selected coronary vessel, the cross-sectional image being generated using one of the set of CT images of the selected coronary vessel. Each of the locations along the at least one SMPR image is associated with one of the CT images in the set of CT images, such that selection of a particular location along the coronary vessel in the at least one SMPR image causes the associated CT image to be displayed in the cross-sectional image of the third panel. In one example, the cross-sectional image may be displayed in panel 606 as shown in FIG. 6A. At block 1035, the process 1000 may receive a second input to the user interface indicating a first location along the selected coronary vessel in the at least one SMPR image. In one example, the user may use a pointing device to select a different portion of the vessel shown in the SMPR image of panel 604. At block 1030, in response to the second input, the process 1000 displays the associated CT scan associated with the cross-sectional image in the third panel, panel 606. That is, the cross-sectional image corresponding to the first input is replaced with the cross-sectional image corresponding to the second input for the SMPR image.

[0426] Normalization Device In some examples, medical images processed and / or analyzed as described throughout this specification can be normalized using a normalization device. As described in more detail in this section, the normalization device may comprise a device that includes multiple samples of known materials that can be placed in the field of view of the medical image to provide an image of the known materials that can serve as a basis for normalizing the medical image. In some examples, the normalization device allows for direct in-image comparison of patient tissue and / or other materials (e.g., plaque) in the image to the known materials in the normalization device.

[0427] As briefly mentioned above, in some instances, a medical imaging scanner may produce images with different scalable radiodensities for the same object. This may vary, for example, depending on the type of medical imaging scanner or device used, as well as the scanning parameters and / or environment on the particular day and / or time the scan was performed. As a result, even when two different scans of the same subject are performed, the resulting medical image may differ in brightness and / or darkness, which may compromise the accuracy of the analysis results processed from that image. To account for such differences, in some embodiments, a normalization device comprising one or more known samples of a known substance may be scanned with the subject, and the resulting image of the one or more known elements may be used as a reference for translating, transforming, and / or normalizing the resulting image.

[0428] Normalization of the medical images being analyzed can be beneficial for several reasons. For example, medical images can be captured under a wide variety of conditions, all of which can affect the resulting medical images. In an example where the medical imaging device comprises a CT scanner, a number of different variables can affect the resulting images. For example, variable image acquisition parameters can affect the resulting images. Variable image acquisition parameters can include one or more of kilovoltages (kV), kilovoltage peak (kVp), milliamps (mA), or gating methods, among others. In some embodiments, gating methods can include predictive axial triggering, retrospective ECG helical gating, and fast pitch helical, among others. Varying any of these parameters can result in slight differences in the resulting medical images, even when the same subject is scanned.

[0429] In addition, the type of reconstruction used to prepare the image after scanning can result in differences in medical images. Example types of reconstruction can include iterative reconstruction, non-iterative reconstruction, machine learning-based reconstruction, and other types of physics-based reconstruction, among others. Figures 11A-11D show different images reconstructed using different reconstruction techniques. In particular, Figure 11A shows a CT image reconstructed using filtered backprojection, and Figure 11B shows the same CT image reconstructed using iterative reconstruction. As shown, the two images look slightly different. The normalization device described below can be used to help take these differences into account by providing a way to normalize between the two. Figure 11C shows a CT image reconstructed by using iterative reconstruction, and Figure 11D shows the same image reconstructed using machine learning. Again, it can be seen that the images contain slight differences, and the normalization device described herein can be advantageously useful to normalize the images to take the differences between the two into account.

[0430] As another example, various types of image capture technologies can be used to capture medical images. In an example where the medical imaging device comprises a CT scanner, such image capture technologies may include dual source scanners, single source scanners, dual energy, monochromatic energy, spectral CT, photon counting, and different detector materials, among others. As mentioned above, images captured using different parameters may appear slightly different even when scanning the same subject. In addition to CT scanners, other types of medical imaging devices can also be used to capture medical images. These can include, for example, x-ray, ul...

Claims

1. 1. A computer-implemented method for assisting in treatment decisions for a subject with coronary artery disease (CAD) based on a risk assessment of the subject, comprising: determining, by a computer system, a CAD risk factor level for the subject; accessing, by a computer system, medical images of a subject including one or more coronary arteries; analyzing, by a computer system, medical images of the subject to determine a CAD risk assessment for the subject based at least in part on quantitative phenotyping of vascular morphology and atherosclerosis, including analysis of one or more quantified plaque parameters; determining, by a computer system, a correlation between CAD risk factor levels and CAD risk assessments; determining, by the computer system, a subject-specific target for the CAD risk factor level, the target being configured to be used to determine an individualized treatment for the subject based on the determined correlation between the CAD risk factor level and the CAD risk assessment; A computer-implemented method, wherein the computer system includes a computer processor and an electronic storage medium.

2. 2. The computer-implemented method of claim 1, wherein the CAD risk factor levels include one or more of cholesterol levels, low-density lipoprotein (LDL) cholesterol levels, high-density lipoprotein (HDL) cholesterol levels, cholesterol particle size and softness, inflammation levels, glycosylated hemoglobin, and blood pressure.

3. The computer-implemented method of claim 1 , wherein quantitative phenotyping of atherosclerosis is performed based at least in part on an analysis of density values ​​of one or more pixels of the medical image corresponding to plaque.

4. The computer-implemented method of claim 3 , wherein the density values ​​comprise radiological concentration values.

5. 2. The computer-implemented method of claim 1, wherein the quantified plaque parameters include at least plaque volume, the plaque volume including one or more of total plaque volume, calcified plaque volume, non-calcified plaque volume, or low-density non-calcified plaque volume.

6. 2. The computer-implemented method of claim 1, wherein the quantified plaque parameters include at least plaque composition, the plaque composition including one or more of calcified plaque, non-calcified plaque, or low-density non-calcified plaque.

7. 7. The computer-implemented method of claim 6, wherein one or more of calcified plaque, non-calcified plaque, or low-density non-calcified plaque is identified based at least in part on radiodensity values ​​of one or more pixels of the medical image corresponding to the plaque.

8. 8. The computer-implemented method of claim 7, wherein the calcified plaque comprises one or more pixels of the medical image having a radiodensity value of about 351 to about 2500 Hounsfield units, the non-calcified plaque comprises one or more pixels of the medical image having a radiodensity value of about 31 to about 250 Hounsfield units, and the low-density non-calcified plaque comprises one or more pixels of the medical image having a radiodensity value of about -189 to about 30 Hounsfield units.

9. 10. The computer-implemented method of claim 1, wherein the quantified plaque parameters include at least plaque progression, and wherein determining plaque progression is performed by the computer system accessing one or more sequential medical images of the subject, the sequential medical images including one or more coronary arteries, and analyzing the one or more sequential medical images of the subject to determine plaque progression based at least in part on sequential changes in plaque volume.

10. 10. The computer-implemented method of claim 9, wherein the continuous change in plaque volume is based on one or more of total plaque volume, calcified plaque volume, non-calcified plaque volume, or low-density non-calcified plaque volume.

11. 2. The computer-implemented method of claim 1, wherein the vascular morphology comprises one or more of absolute minimum luminal diameter or area, luminal diameter, luminal cross-sectional area, vascular volume, luminal volume, arterial remodeling, vascular or luminal geometry, or vascular or luminal curvature.

12. The computer-implemented method of claim 1 , wherein the correlation between CAD risk factor levels and CAD risk assessment is determined based at least in part on a machine learning algorithm.

13. The computer-implemented method of claim 1 , wherein the medical images include computed tomography (CT) images.

14. 10. The computer-implemented method of claim 1, wherein the medical images are obtained using imaging techniques including one or more of CT, X-ray, ultrasound, echocardiography, intravascular ultrasound (IVUS), MR imaging, optical coherence tomography (OCT), nuclear medicine imaging, positron emission tomography (PET), single photon emission computed tomography (SPECT), or near-field infrared spectroscopy (NIRS).

15. The computer-implemented method of claim 1 , wherein treating cardiovascular disease comprises medical intervention, drug therapy, or lifestyle intervention.

16. 10. The computer-implemented method of claim 1, further comprising: accessing, by the computer system, a second medical image of the subject obtained at a time later than the first medical image; analyzing, by the computer system, the second medical image of the subject to determine a CAD risk assessment for the subject at the subsequent time point; and updating, by the computer system, a subject-specific goal for CAD risk factor levels based at least in part on the subject's CAD risk assessment at the subsequent time point; wherein the updated subject-specific goal for CAD risk factor levels is configured to be used to modify or maintain an individualized treatment for the subject.

17. 1. A system for assisting a subject with coronary artery disease (CAD) in making a treatment decision based on a CAD risk assessment derived from non-invasive medical image analysis, comprising: one or more computer-readable storage devices configured to store a plurality of computer-executable instructions; one or more hardware computer processors in communication with one or more computer-readable storage devices and configured to execute a plurality of computer-executable instructions; determining a CAD risk factor level for the subject by executing a plurality of computer-executable instructions; accessing a medical image of a subject including one or more coronary arteries; analyzing the subject's medical images to determine a CAD risk assessment for the subject based on a quantitative phenotyping of atherosclerosis, the quantitative phenotyping including analysis of one or more quantified plaque parameters; determining a correlation between CAD risk factor levels and CAD risk assessment; and determining a subject-specific target for the CAD risk factor level, the target being configured for use in determining an individualized treatment for the subject based on the determined correlation between the CAD risk factor level and the CAD risk assessment.

18. 18. The system of claim 17, wherein the CAD risk factor levels include one or more of cholesterol levels, low-density lipoprotein (LDL) cholesterol levels, high-density lipoprotein (HDL) cholesterol levels, cholesterol particle size and softness, inflammation levels, glycosylated hemoglobin, and blood pressure.

19. 20. The system of claim 17, wherein quantitative phenotyping of atherosclerosis is performed based at least in part on an analysis of density values ​​of one or more pixels of the medical image corresponding to plaque.

20. 20. The system of claim 19, wherein the density values ​​comprise radiological concentration values.