Systems and methods for detecting microcalcification activity - Patents.com

Abstract

A system and method for predicting microcalcification activity in a blood vessel, including either an artery or a vein, comprising: (a) measuring patient data including one or more of the following: the presence and / or amount of coronary plaque or visible disease markers in a vascular tissue sample; the presence and / or amount of healthy tissue in the vascular tissue sample; one or more features defining an abnormal hemodynamic environment in the blood vessel; one or more geometric features associated with vascular remodeling and affecting vascular hemodynamics, and / or one or more material properties affecting vascular hemodynamics; and (b) calculating the microcalcification activity in the blood vessel as a function of the measurements obtained in step (a).

Description

[Technical field]

[0001]

[0001] The present invention generally relates to a method for the treatment of vascular 18 In the field of estimating / predicting F-NaF uptake, in particular: 18 in vascular tissue without F-NaF PET imaging 18 It is directed to estimating / predicting F-NaF uptake and is described below with reference to this application, although it is understood that the invention is not limited to this particular field of use. [Background technology]

[0002]

[0002] Any discussion of background art throughout this specification should in no way be deemed to be an admission that such background art is prior art and that such background art is widely known or forms part of the common general knowledge in the field in Australia or throughout the world.

[0003]

[0003] All references cited in this specification, including any patents or patent applications, are incorporated herein by reference. No admission is made that any reference constitutes prior art. The discussion of the references states what their authors assert, and the applicants reserve the right to challenge the accuracy and pertinence of the cited documents. Although a number of prior art publications are referenced in this specification, it is expressly understood that this reference is not an admission that any of these documents form part of the common general knowledge in the art in Australia or any other country.

[0004]

[0004] Although this discussion relates specifically to imaging and analysis of microcalcification activity in the coronary arteries causing coronary heart disease, the disclosure herein is not limited to the coronary arteries and is readily applicable to a patient's systemic vascular system, including, in particular, at least the patient's carotid arteries, cerebral arteries, aorta, peripheral arteries, or any artery or vein of the vascular system.

[0005]

[0005] Coronary heart disease (CHD) is the leading cause of death in the world. In 2015, CHD affected 110 million people and killed 8.9 million people. This represents 16% of all deaths, making it the leading cause of death worldwide. Coronary heart disease is also the single leading cause of death in Australia (12% of all deaths, one death every 27 minutes).

[0006]

[0006] In 2016, the American Heart Association (HCA) reported that 15.5 million people aged 20 years or older in the United States have CHD. The reported prevalence increases with age in both women and men.

[0007] Approximately 25-30% of patients hospitalized with a heart attack due to CHD will die or be re-admitted at least once for a further clinical event within 3 years, resulting in a significant health burden. Recurrent events occur despite the routine use of several proven risk mitigation strategies, including inpatient coronary angiography and revascularization, statins, dual antiplatelet therapy, beta-blockers, ACE inhibitors, and lifestyle advice. For example, novel preventive therapies targeting inflammatory processes associated with plaque rupture are being evaluated, but are costly and may have undesirable side effects. It is widely recognized that there is an unmet need for improved risk stratification to better apply these therapies only to those patients who will benefit most.

[0008]

[0008] If a patient is suspected of having CHD, the treating cardiologist will usually perform a percutaneous coronary intervention (PCI). PCI involves inserting a catheter into the heart via the wrist or groin. Injecting a contrast agent under x-ray imaging highlights the blood flow and reveals the narrowing of the arteries. This part of the procedure is called a coronary angiogram and can be recorded for subsequent analysis. The cardiologist then guides another catheter towards the blockage and opens the blockage with a balloon or places a stent in the blocked area. PCI procedures help relieve the symptoms of coronary artery disease and reduce damage to the heart after or during a heart attack. The global PCI market was US$10 billion in 2017 and is expected to exceed US$15 billion by 2023.

[0009]

[0009] During coronary angiography and prior to angioplasty (balloon therapy) or stenting, important information can be obtained regarding the pressure difference along the diseased artery due to blockage, which informs the decision to use a stent or not. This is called fractional flow reserve (FFR) and measures the pressure difference across a coronary artery stenosis (usually due to atherosclerosis) and determines the likelihood that the stenosis will impede oxygen delivery to the heart muscle (myocardial ischemia). FFR has become the standard of care for assessing the physiological significance of coronary artery disease (CAD). When FFR is used to guide percutaneous coronary intervention (PCI), clinical outcomes are improved, fewer stents are deployed, and costs are reduced.

[0010]

[0010] However, even in countries where FFR is most frequently used, it is used in less than 10% of PCI procedures and much less in diagnostic cases; therefore, despite these advantages, clinical uptake remains extremely low. This is due to a combination of factors related to practicality, time and cost. Using computational fluid dynamics (CFD) to calculate a "virtual" FFR (vFFR) from coronary artery angiograms (CAG) offers the advantages of physiologically guided PCI without the drawbacks that limit invasive techniques.

[0011]

[0011] vFFR can be calculated based on three-dimensional (3D) reconstruction of coronary anatomy from coronary CT angiograms (images) using CFD modeling. Recently, the FDA approved clinical implementation of FFR derived from computational modeling. Optimized methods of determining vFFR (e.g., 3D-QCA derived vFFR) provide results within 4 minutes or near real-time. Although early results have been promising, the accuracy of vFFR calculation is limited by the accuracy with which the models represent coronary artery and lesion geometry (imaging and reconstruction) and physiological parameters (adjustment of boundary conditions) on an individual patient basis. The last major barrier to a reliable vFFR tool is the application of patient-specific adjustment strategies to represent hyperemic flow or myocardial resistance.

[0012]

[0012] Better imaging systems can be used to establish optimized treatment plans for stent implantation, especially in patients with comorbid conditions that were at higher risk for cardiac events. A recent network meta-analysis clearly demonstrated the superiority of intravascular ultrasound (IVUS) and / or optical coherence tomography (OCT) over coronary angiography guidance (Buccheri et al., 2017). In particular, since angiography is known to have limitations in assessing vessel size and plaque burden, lesion calcium and eccentricity, stent expansion, and geographical errors and complications, IVUS and OCT are now used to answer questions that arise during routine PCI. The use of OCT is rapidly replacing older technologies such as IVUS due to its 10 times higher image resolution and faster image acquisition time; for example, in Japan, OCT is used in approximately 80% of all PCIs. Therefore, experts believe that OCT analysis will become an important tool and that automated image analysis will be urgently needed.

[0013]

[0013] One of the main advantages of using OCT over other imaging modalities is the ability to measure thin-capped fibrous atheroma (TCFA), a rim of fibrous tissue that separates the necrotic core of the plaque from the lumen of the artery. Rupture of a TFCA means that the contents of the necrotic core will spill into the bloodstream and cause downstream blockage. The most dangerous TCFA thickness is less than 65 μm, which can only be measured using OCT. Although TCFA thickness is one measure of risk, the unpredictability of coronary artery disease is due to the behavior of unstable and stable plaques, and currently there is no way to assess plaque stability.

[0014] In addition to TCFA, many other biomarkers of plaque stability can be visualized and quantified on OCT. Characteristics of current clinical interest include: (1) Lipid region; (2) surface calcium; (3) deep calcium; (4) plaque-free walls; (5) Thrombus; (6) Macrophages; (7) Microchannel; (8) Cholesterol crystals; and of course, (9) Thin-capsulated fibrous atheroma.

[0015]

[0015] Each of these contributes to the risk of plaque rupture causing heart attacks and potential death. Although the importance of these features is recognized, currently they must be manually annotated and quantified on each OCT image (of which there are typically about 500 per artery segment). This is not only extremely time consuming, but also introduces user-to-user variability.

[0016]

[0016] In addition to image-based biomarkers of plaque stability, growing evidence suggests that biomechanical aspects are important for plaque assessment. From a biomechanical perspective, two forces are usually considered: shear stress and structural stress.

[0017]

[0017] Shear stress is the frictional force acting on the inside of a blood vessel (i.e., either an artery or a vein) wall and on plaque by flowing blood. Computational fluid dynamics (CFD) is the primary method used to calculate the shear stress acting on the inner wall of a patient's artery or vein. A certain level of shear stress is required for normal physiological function, and low shear stress is a well-established predictor of plaque progression and future clinical events. Currently, there are no commercially available tools to provide clinicians with usable vascular shear stress information.

[0018]

[0018] Structural stresses acting on blood vessels and plaques by blood pressure are also of great clinical value. If the structural stress exceeds the structural strength of the tissue, the plaque will rupture. This has been the focus of major research activity for many years, and significant progress has been made. However, as with shear stress, there are currently no commercially available tools available to clinicians that can provide these important data.

[0019]

[0019] Currently, there is no semi-automated or automated method to analyze transluminal OCT image data or to calculate the critical biomechanical forces that destabilize plaque and cause life-threatening plaque rupture (heart attack). Additionally, clinicians still need intravascular imaging and software pre-PCI treatment planning tools that can comprehensively select optimal stent size, length and placement to minimize further damage to the artery.

[0020]

[0020] Detected by Positron Emission Tomography (PET) 18 F-Sodium fluoride ( 18 F-NaF) uptake has been associated with high-risk coronary plaque characteristics and future clinical events (Joshi et al., 2014). 18 F-fluoride positron emission tomography for identification of ruptured and high risk coronary atherosclerotic plaques;a prospective clinical trial.Lancet,383,705-13;Lee et al., 2017, Clinical Relevance of (18)F-Sodium Fluoride Positron-Emission Tomography in Noninvasive Identification of High-Risk Plaque in Patients with Coronary Artery Disease.Circ Cardiovasc Imaging,10). 18 Areas of F-NaF uptake indicate microcalcification activity in the arterial wall, and these areas may develop into macrocalcification in the future. 18 F-NaF PET has been shown to provide a strong independent prediction of fatal or non-fatal myocardial infarction (Kwiecinski et al., 2020, Coronary 18F-Sodium Fluoride Uptake Predicts Outcomes in Patients With Coronary Artery Disease. J Am Coll Cardiol, 75, 3061-74). Kwiecinski et al. 18 The uptake of F-NaF in blood vessels 18 Based on the amount and intensity of F-NaF PET activity, the data were converted into an equivalent measure called coronary microcalcification activity (CMA), which represents overall disease activity within blood vessels and therefore serves as an early indicator of future adverse events. 18 Detecting, visualizing, and quantifying F-NaF uptake is of great clinical relevance. However, this imaging technique is expensive, available only in specialized centers, and requires significant technical expertise to analyze the images. Furthermore, patients are exposed to additional radiation, and some patients do not tolerate imaging radiotracers.

[0021]

[0021] 18 The F-NaF tracer, along with other blood-borne particles, is transported through the bloodstream and binds to early active vascular microcalcification (Irkle et al., 2015, Identifying active vascular microcalcification by 18 F-sodium fluoride positron emission tomography.Nature Communications,6,7495). 18The F-NaF tracer is therefore influenced by hemodynamics (transport to the plaque site) and the distribution of plaque within the coronary artery. Proliferating / active coronary plaques are thought to be preferential binding sites for NaF due to the occurrence of microcalcification activity, which is caused by cell death and inflammation (Chen and Dilsizian, 2013, Targeted PET / CT Imaging of Vulnerable Atherosclerotic Plaques: Microcalcification with Sodium Fluoride and Inflammation with Fluorodeoxyglucose. Current Cardiology Reports, 15, 364).

[0022]

[0022] It has been shown that coronary blood supply depends on geometric measurements such as vessel diameter / caliber, vessel length and myocardial muscle mass (Zamir et al., 1992, Relation between diameter and flow in major branches of the arch of the aorta. J Biomech, 25, 1303-10; Choy and Kassab, 2008, Scaling of Myocardial Mass to Flow and Morphometry of Coronary Arteries. Journal of Applied Physiology (Bethesda, Md.: 1985), 104, 1281-1286). Coronary plaque propagation changes the local vessel caliber and shape through remodeling (Libby and Theroux, 2005, Pathophysiology of coronary artery disease. Circulation, 111, 3481-8). This shape change is often described by geometric measures such as area and eccentricity (Hausmann et al., 1994, Lumen and plaque shape in atherosclerotic coronary arteries assessed by in vivo intracoronary ultrasound. Am J Cardiol, 74, 857-63) and is associated with increased structural stress within the plaque (Costopoulos et al., 2017, Plaque Rupture in Coronary Atherosclerosis Is Associated with Increased Plaque Structural Stress. JACC Cardiovasc Imaging, 10, 1472-1483).Plaque characteristics are associated with both high and low endothelial wall shear stress (WSS), which is the fluid friction force acting on the endothelium (Koskinas et al., 2009, The role of low endothelial shear stress in the conversion of atherosclerotic lesions from stable to unstable plaque. Curr Opin Cardiol, 24, 580-90.). The value of WSS at a particular location depends on many factors, including blood supply, blood properties (e.g., viscosity), vessel inner diameter and lumen geometry (e.g., eccentricity, curvature (Myers et al., 2001, Factors Influencing Blood Flow Patterns in the Human Right Coronary Artery. Annals of Biomedical Engineering, 29, 109-120.)). Low WSS stimulates the atheromatous phenotype, promoting arterial inflammation and the expression of adhesion proteins and chemokines that cooperate to trap leukocytes from the bloodstream into the vessels (Malek et al., 1999, Hemodynamic shear stress and its role in atherosclerosis. Jama, 282, 2035-42; Lawrence et al., 1987, Effect of flow on polymorphonuclear leukocyte / endothelial cell adhesion. Blood, 70, 1284-90; and Gijsen et al., 2019, Expert recommendations on the assessment of wall shear stress in human coronary arteries: existing methodologies, technical considerations, and clinical applications. European Heart Journal, 40, 3421-3433.).Furthermore, low WSS exists in regions of recirculation flow and low near-wall velocities, which aid in the aggregation of blood-borne particles near the endothelium (Gijsen et al., 2019. Expert recommendations on the assessment of wall shear stress in human coronary arteries: existing methodologies, technical considerations, and clinical applications. European Heart Journal, 40, 3421-3433.).These hemodynamic mechanisms and associated endothelial dysfunction help explain why plaque progression occurs in areas of low shear stress (Chatzizisis et al., 2008, Prediction of the Localization of High-Risk Coronary Atherosclerotic Plaques on the Basis of Low Endothelial Shear Stress. Circulation, 117, 993-1002; Stone et al., 2012, Prediction of Progression of Coronary Artery Disease and Clinical Outcomes Using Vascular Profiling of Endothelial Shear Stress and Arterial Plaque Characteristics: The PREDICTION Study. Circulation; Kumar et al., 2018, Low Coronary Wall Shear Stress Is Associated with Severe Endothelial Dysfunction in Patients with Nonobstructive Coronary Artery Disease. JACC Cardiovasc Interv,11,207-2080;Yamamoto et al., 2017, Low Endothelial Shear Stress Predicts Evolution to High-Risk Coronary Plaque Phenotype in the Future.Circulation:Cardiovascular Interventions,10,e005455; and Bourantas et al., 2019, Implications of the local hemodynamic forces on the phenotype of coronary plaques.Heart,heartjnl-2018-314086).

[0023]

[0023] We consider the same environment to facilitate detection of microcalcification activity through local NaF tracer retention, infiltration and binding.

[0024]

[0024] Attempts to use both shear and structural stresses clinically are currently limited by significant difficulties, such as automated image analysis, computation time, and a lack of important boundary condition data for simulation. Solutions to these problems have not been described so far. Furthermore, 18 Quantitative data on F-NaF uptake (i.e., active microcalcifications) show clinical promise, but are expensive and inaccessible. 18 F-NaF PET hardware, 18 There is a strong need to overcome the current requirement for nuclear medicine and image analysis expertise to access and interpret F-NaF data, problems of excessive radiation and patient intolerance to radiotracers.

[0025]

[0025] It is against this background that the present invention has been developed. In particular, the present invention aims to overcome or at least ameliorate one or more of the shortcomings of the prior art discussed above, or to provide consumers with a useful or commercial choice. Summary of the Invention

[0026] It is an object of the present invention to overcome or ameliorate at least one or more of the disadvantages of the prior art, or to provide a useful alternative.

[0027]

[0027] One embodiment provides a computer program product for performing the methods described herein.

[0028]

[0028] One embodiment provides a non-transitive carrier medium for carrying computer executable code that, when executed on a processor, causes the processor to perform a method described herein.

[0029]

[0029] One embodiment provides a system configured to perform the methods described herein.

[0030]

[0030] The present invention relates to a principle of general application in that it provides a method for measuring microcalcification activity in arteries (preferably coronary arteries) using anatomical, image, plaque and blood flow / hemodynamic measurements. In particular, the present invention provides a means to accommodate measurements of regions of interest and lesions within healthy and / or diseased vasculature and to identify / quantify / distinguish between these conditions. The list of possible inputs and outputs for use with the method is extensive, and microcalcification prediction is a result that may be derived / possibly intended from the method.

[0031] According to the present invention, we provide a method for measuring microcalcification activity in blood vessels, preferably coronary arteries.

[0032] According to a first aspect of the present invention, there is provided a method for predicting microcalcification activity in a blood vessel. The blood vessel may comprise either an artery or a vein. The method may comprise the step (a) of measuring one or more of: the presence and / or amount of coronary plaque or visible disease markers in a blood vessel tissue sample; and / or the presence and / or amount of healthy tissue in a blood vessel tissue sample; and / or one or more features defining an abnormal hemodynamic environment in the blood vessel; and / or one or more geometric features associated with vascular remodeling and affecting vascular hemodynamics; and / or one or more material properties affecting vascular hemodynamics. The method may further comprise the step (b) of calculating the microcalcification activity in the blood vessel as a function of the measurements obtained in step (a).

[0033] According to a particular configuration of the first aspect, there is provided a method of predicting microcalcification activity in a blood vessel, including either an artery or a vein, the method comprising: (a) measuring one or more of the following: (i) the presence and / or amount of coronary plaque or visible disease markers in a vascular tissue sample; and / or (ii) the presence and / or amount of healthy tissue within the vascular tissue sample; and / or (iii) one or more characteristics defining an abnormal hemodynamic environment in a blood vessel; and / or (iv) one or more geometric features associated with vascular remodeling that affect hemodynamics within the vessel; and / or (v) measuring one or more of the one or more material properties that affect vascular hemodynamics; (b) calculating microcalcification activity in the blood vessel as a function of the measurements obtained in step (a).

[0034] According to a second aspect of the present invention, there is provided a method for predicting microcalcification activity in blood vessels. The method may comprise the steps of obtaining training data. The training data may consist of general patient data including data for a plurality of patients, a plurality of data; and microcalcification activity data (μCA) from a plurality of patients at one or more anatomical locations. The method includes fitting a multivariate function / model by calculating a function from input patient data and microcalcification activity data, and calculating a function from the input training data set [A Tr , B Tr , C Tr , D Tr , ...] obtained in the same way as new , b new , c new , d new ...], said function being: The file is TIFF2024529405000002.tif6170.

[0035] The method evaluates the multivariate model by evaluating a pre-fitted function to obtain a set of estimates of microcalcification activity (μCA) for a new set of function inputs. Est The method may further include obtaining a set of estimates of microcalcification activity (μEst ) with a set of known / matched values ​​of microcalcification activity data (μCA Te ) using an error function E to calculate an error estimate: TIFF2024529405000003.tif6170

[0036]

[0036] The method may further comprise the step of checking the error estimates to assess the suitability of the model.

[0037] According to a particular configuration of the second aspect, there is provided a method of predicting microcalcification activity in a blood vessel, the method comprising: (i) general patient data, including data relating to a plurality of patients, and a plurality of data; (ii) obtaining training data comprised of microcalcification activity data (μCA) from a plurality of patients at one or more anatomical locations; A function is calculated from the input patient data and the microcalcification activity data to fit a multivariate function / model, and the input training data set [A Tr , B Tr , C Tr , D Tr , ...] obtained in the same way as new , b new , c new , d new ....], said function being: TIFF2024529405000004.tif7170, steps; Evaluate the multivariate model by evaluating the pre-fitted functions and the set of estimates of microcalcification activity (μCA) for a new set of function inputs. Est ) and; A set of estimates of microcalcification activity (μCA Est) with a set of known / matched values ​​of microcalcification activity data (μCA Te ) using an error function Ef to calculate an error estimate, TIFF2024529405000005.tif7170, steps; Checking the error estimates to assess the suitability of the model.

[0038]

[0038] According to a third aspect of the present invention, there is provided a method for predicting microcalcification activity in a blood vessel. The method may include receiving training data associated with one or more of the following: a presence and / or amount of coronary plaque or visible disease markers in a blood vessel tissue sample; and / or a presence and / or amount of healthy tissue in a blood vessel tissue sample; and / or one or more features defining an abnormal hemodynamic environment in the blood vessel; and / or one or more geometric features associated with vascular remodeling and affecting hemodynamics in the blood vessel; and / or one or more material properties affecting vascular hemodynamics. The method may further include determining one or more training features based on the training data values. The method may further include determining one or more training labels associated with the one or more training features. The method may further include constructing a predictive model using a computer to determine microcalcification activity in the blood vessel. The step of constructing the predictive model may include inputting the one or more training features and the one or more training labels associated with the one or more training features into a machine learning algorithm. The step of constructing the predictive model may further include determining, from the machine learning algorithm, a predictive model for receiving new data associated with the vessel; and determining a predictive label based on the new data.

[0039] According to a particular configuration of the third aspect, there is provided a method of predicting microcalcification activity in a blood vessel, the method comprising: (i) the presence and / or amount of coronary plaque or visible disease markers in a vascular tissue sample; and / or (ii) the presence and / or amount of healthy tissue within the vascular tissue sample; and / or (iii) one or more characteristics defining an abnormal hemodynamic environment in a blood vessel; and / or (iv) one or more geometric features associated with vascular remodeling that affect hemodynamics within the vessel; and / or (v) receiving training data associated with one or more of the one or more material properties that affect vascular hemodynamics; determining one or more training features based on the training data values; determining one or more training labels associated with the one or more training features; Using a computer to construct a predictive model for determining microcalcification activity in a blood vessel, the step of constructing the predictive model comprising: inputting one or more training features and one or more training labels associated with the one or more training features into a machine learning algorithm; Building the predictive model includes determining, from the machine learning algorithm, a predictive model for receiving new data associated with the vessel and determining a predictive label based on the new data.

[0040] According to a fourth aspect of the present invention, there is provided a computer-implemented method for measuring microcalcification activity in a blood vessel. The computer-implemented method may include the steps of: (a) measuring one or more of the presence and / or amount of coronary plaque or visible disease markers in a blood vessel tissue sample; and / or the presence and / or amount of healthy tissue in a blood vessel tissue sample; and / or one or more features defining an abnormal hemodynamic environment in the blood vessel; and / or one or more geometric features associated with vascular remodeling and affecting hemodynamics in the blood vessel; and / or one or more material properties affecting vascular hemodynamics. The computer-implemented method may further include the step of (b) calculating the microcalcification activity in the blood vessel as a function of the measurements obtained in step (a) using the trained machine learning model.

[0041] According to a particular configuration of the fourth aspect, there is provided a computer-implemented method for measuring microcalcification activity in a blood vessel, the method comprising: (a) measuring the steps of: (i) the presence and / or amount of coronary plaque or visible disease markers in a vascular tissue sample; and / or (ii) the presence and / or amount of healthy tissue within the vascular tissue sample; and / or (iii) one or more characteristics defining an abnormal hemodynamic environment in a blood vessel; and / or (iv) one or more geometric features associated with vascular remodeling that affect hemodynamics within the vessel; and / or (v) measuring one or more of the one or more material properties that affect vascular hemodynamics; (b) using the trained machine learning model to calculate microcalcification activity in the blood vessel as a function of the measurements obtained in step (a).

[0042]

[0042] The first machine learning model may include a first trained regression model.

[0043]

[0043] The blood vessel may be one or more of a coronary artery, a carotid artery, a cerebral artery, an aorta, a peripheral artery or a vein.

[0044] According to a fifth aspect of the invention there is provided a method of providing information for predicting 18F-NAF uptake in vascular tissue of a patient. The method may comprise the step of measuring, using image processing means on the patient image data, vascular biomarkers indicative of the presence and / or amount of coronary plaque or visible disease markers in the vascular tissue associated with the progression of cardiovascular disease. The method may comprise the further step of calculating, using a processor, microcalcification activity in the vascular tissue as a function of the measurements.

[0045]

[0045] According to a particular configuration of the fifth aspect, there is provided a method for providing information for predicting uptake of 18F-NAF in a patient's vascular tissue, the method comprising the steps of: using image processing means on patient image data to measure the presence and / or amount of visible disease markers in the vascular tissue associated with coronary plaque or cardiovascular disease progression; and using a processor to calculate microcalcification activity in the vascular tissue as a function of the measurements.

[0046] According to a sixth aspect of the present invention, there is provided a computer system comprising at least one processor and at least one memory device for storing patient data. The stored patient data may be related to the presence and / or amount of coronary plaque or visible disease markers in the vascular tissue sample; and / or the presence and / or amount of healthy tissue in the vascular tissue sample; and / or one or more features defining an abnormal hemodynamic environment in the blood vessel; and / or one or more geometric features associated with vascular remodeling and affecting hemodynamics in the blood vessel; and / or one or more material properties affecting vascular hemodynamics. The at least one processor may be configured to use the trained machine learning model to calculate microcalcification activity in the blood vessel as a function of the patient data.

[0047] According to a particular configuration of the sixth aspect, there is provided a computer system, the computer system comprising: At least one processor; At least one memory device: (i) the presence and / or amount of coronary plaque or visible disease markers in a vascular tissue sample; and / or (ii) the presence and / or amount of healthy tissue within the vascular tissue sample; and / or (iii) one or more characteristics defining an abnormal hemodynamic environment in a blood vessel; and / or (iv) one or more geometric features associated with vascular remodeling that affect hemodynamics within the vessel; and / or (v) at least one memory device that stores patient data related to one or more material properties that affect vascular hemodynamics; The at least one processor is configured to use the trained machine learning model to calculate microcalcification activity in the blood vessel as a function of patient data.

[0048]

[0048] The method of any one of the above aspects may include the step of measuring microcalcification activity in any subject's blood vessels, including coronary arteries, carotid arteries, cerebral arteries, aorta, peripheral arteries or veins.

[0049] According to a further aspect of the present invention there is provided a method of predicting microcalcification activity in a blood vessel comprising: A step of obtaining training data (Tr), the training data (Tr) comprising: General patient data [A] including data for a plurality of patients, and a plurality of data that may further include image data and / or biomechanical data at one or more anatomical locations. Tr , B Tr , C Tr , D Tr ...], and comprising microcalcification activity data (μCA) from a plurality of patients at one or more anatomical locations; Input patient training data and input training dataset A Tr , B Tr , C Tr , D Tr New data a obtained in the same way as ... new , b new , c new , d new and microcalcification activity data from which μCA can be newly estimated / predicted for .... and fitting a multivariate function / model by calculating a function from: TIFF2024529405000006.tif6170, and We evaluate the multivariate model by evaluating a pre-fitted function f on the training set [A Te , B Te , C Te , D Te ...] for a set of new function inputs, which is a set of new general patient test data (Te) that was not included in Est ) to obtain a set of estimates of TIFF2024529405000007.tif6170, and calculating an error estimate using an error function Ef by comparing the set of estimated values ​​(Est) of microcalcification activity (μCAEst) with a set of known / corresponding values ​​of microcalcification activity data (μCATe) derived from a corresponding set of data collected for the same patient and used to generate the function inputs [ATe, BTe, CTe, DTe...]; TIFF2024529405000008.tif7170, and Checking the error estimates / assessing the suitability of the model If the error estimates meet a set of desired criteria such as required accuracy, precision and sensitivity for the input data where the errors are small and the results are statistically significant, the model is deemed fit for purpose and may be used to predict microcalcification activity.

[0050] According to a further aspect of the present invention, there is provided a non-transitory computer readable medium storing computer program instructions for measuring microcalcification activity in a blood vessel, preferably a coronary artery, the computer program instructions, when executed by a processor, causing the processor to: (a) measuring, (i) the presence and amount of plaque or visible disease markers in the vascular tissue sample; (ii) the presence and / or amount of healthy tissue within the vascular tissue sample; (iii) one or more characteristics defining an abnormal hemodynamic environment within the blood vessel; (iv) one or more geometric features associated with vascular remodeling and affecting hemodynamics within the blood vessel; and / or (v) measuring one or more of the material properties that affect vascular hemodynamics; (b) calculating microcalcification activity in the blood vessel as a function of the measurements obtained in step (a).

[0051] According to a further aspect of the present invention, there is provided a non-transitory computer readable medium storing computer program instructions for measuring microcalcification activity in a blood vessel, preferably a coronary artery, the computer program instructions, when executed by a processor, causing the processor to: [A Tr , B Tr , C Tr , D Tr , ...] and general patient data, obtaining training data comprised of microcalcification activity data (μCA) from a plurality of patients at one or more anatomical locations; Fitting a multivariate function / model by calculating a function from the input patient data and microcalcification activity data that can estimate / predict μCA for new data obtained in the same way as the input training data, said function comprising: TIFF2024529405000009.tif7170, and A set of estimates of microcalcification activity (μCA) for a new set of function inputs, which is a new set of general patient test data that was not included in the training set. Est ) by evaluating the pre-fitted function f, Te , B Te , C Te , D Te ...] includes data relating to a plurality of patients and a plurality of data, which may further include image data and / or biomechanical data at one or more anatomical locations; TIFF2024529405000010.tif6170, and The error function Ef is used to calculate the set of estimates of microcalcification activity (μ Est ), (μCA Te ) to a set of known / corresponding values ​​of TIFF2024529405000011.tif7170, and Checking the error estimates / assessing the suitability of the model. If the error estimates meet a set of desired criteria, such as the required accuracy, precision and sensitivity to the input data, the model is deemed fit for purpose and may be used to predict microcalcification activity.

[0052] According to a further aspect of the present invention, there is provided a non-transitory computer readable medium storing computer program instructions for measuring microcalcification activity in an artery of a patient, the computer program instructions, when executed by a processor, causing the processor to: receiving a multivariate function / model including a function calculated from a training model of input patient data and microcalcification activity CA data capable of estimating / predicting μCA; receiving general patient data for patients that match the input patient data for the training model; The value of microcalcification activity (μCA) in the patient's arteriesEst and calculating

[0053]

[0053] The method of any one of the above aspects may further include one or more of the following features in any combination.

[0054]

[0054] The method may include measuring microcalcification activity in any subject's blood vessels, including coronary arteries, carotid arteries, cerebral arteries, aorta, peripheral arteries or veins.

[0055]

[0055] The patient data and / or training data may include biomarker data relating to one or more characteristics of clinical objects including, but not limited to, lipid regions; surface calcium; deep calcium; plaque-free walls; thrombi; macrophages; microchannels; cholesterol crystals; or thin-capped fibrous atheroma associated with one or more blood vessels of the patient.

[0056]

[0056] The patient data and / or training data may include one or more image data, including but not limited to OCT, angiography, computed tomography (CT); CT angiography image data. The image data may be referenced to a common frame of reference or a common coordinate system. The image data may be transformed to the common frame of reference. The image data may provide a complete representation of the arterial tree of the imaged patient. The representation may be a two-dimensional and / or three-dimensional image representation.

[0057]

[0057] The method may include interpolation of image data between image frames and / or between anatomical landmarks.

[0058]

[0058] The method may include performing a structural simulation at any location of interest along the imaged vessel. The image data may be segmented by a user to identify different regions of plaque and vessel walls within the imaged vessel.

[0059]

[0059] The method may include estimating in vivo material properties based on a tissue stiffness ratio, which may be based on similar in vivo material properties in other regions of the patient's cardiovascular system.

[0060]

[0060] The method may include steps of providing measurements of vascular status, including, but not limited to: intraluminal shear stress; plaque structural stress; plaque characterization; microcalcification activity; virtual stent placement; vessel wall characterization; thin cap measurement; multi-modal imaging; vessel branches; coronary flow reserve; rapid time frames; and VR virtualization.

[0061]

[0061] Microcalcification activity in blood vessels, e.g., arteries such as coronary arteries or veins, may be measured using Positron Emission Tomography (PET), in which any set of measurements may be fitted to a model (via regression or machine learning techniques) to obtain an exact microcalcification activity result.

[0062]

[0062] The presence and / or amount of vascular plaque may be measured based on the measurement of well-established geometric disease markers from intravascular optical coherence tomography (OCT) images, specifically the presence of lipids, calcium and macrophages (bright spots) in the plaque. For example, measurements of the mean lipid arc [°], the mean calcium arc [°], the mean bright spots may be obtained. Additional geometric measurements indicative of disease relate to the vessel diameter, area, volume, arterial wall / layer thickness, tortuosity and eccentricity, as well as all combinations of these measurements. Similarly, these measurements may be obtained by other imaging modalities common in clinical practice.

[0063]

[0063] The presence and / or amount of healthy tissue may be measured based on a measurement of the amount of healthy arterial wall visible using intravascular OCT images, e.g., plaque-free wall (PFW) is inversely correlated with disease. For example, a measurement of the mean arc [°] of the plaque-free wall may be obtained. Similarly, this measurement may be obtained by any other imaging modality typical in clinical practice.

[0064] In the method, measurements of abnormal hemodynamic environment such as blood-borne particle retention or abnormal WSS may be estimated using computational fluid dynamics (CFD) simulations or other methods that can directly estimate wall shear stress (WSS) via imaging techniques. For example, by measuring low shear area (LSA): low WSS area [%] or high shear area (HSA): high WSS area [%] or mean WSS [Pa]. Additional hemodynamic-derived metrics may include, but are not limited to, oscillatory shear index (OSI), relative residence time (RRT), low oscillatory shear (LOS), endothelial activation potential (ECAP), velocity-derived field functions (e.g., vorticity), pressure drop, or the gradient of any of the aforementioned metrics (e.g., gradient of WSS).

[0065] Measurements of geometric features associated with vascular remodeling and its impact on hemodynamics may be obtained from intravascular OCT images (circumferential and eccentric) and coronary computed tomography angiography (CCTA), for example by measuring mean circumference [mm] (using OCT), mean eccentricity (using OCT), arterial wall / layer thickness, and / or ventricular muscle mass [g] (using CT).

[0066]

[0066] Material properties that affect hemodynamics, such as % hematocrit, may be measured during routine blood draws and used to adjust the viscosity model used to calculate the WSS in CFD.

[0067]

[0067] The method may allow for the assessment of treatment options based on one or more subject metrics.

[0068]

[0068] In this method, the imaging diagnostic techniques used in obtaining the measurements of the method may include one or more of: computed tomography (CT); magnetic resonance imaging (MRI); ultrasound; intravenous ultrasound (IVUS); optical coherence tomography (OCT); single photon emission computed tomography (SPECT), PET or NaF PET.

[0069]

[0069] The method may include a fitting process of a multivariate function, for example, parametric or non-parametric regression. Part of the training data may be saved as validation data during the fitting process. The validation data may be used to estimate the prediction error for model selection. Non-parametric regression may include methods such as kernel regression and machine learning support vector machines. Parametric fitting may include using parametric machine learning algorithms or traditional optimization methods to find the minimum of an objective function (e.g., "sum of squared errors"). For non-linear functions, a specific example suitable for use in the method may be a direct search method of multidimensional unconstrained minimization, such as the Nelder-Mead simplex method. Parametric optimization may exploit commonly used functional forms related to biological relationships, such as allometric scaling functions.

[0070]

[0070] Additional objects, advantages and novel features are described in the following description or will become apparent to those skilled in the art upon examination of the drawings and the following detailed description of several non-limiting embodiments.

[0071]

[0071] Notwithstanding any other forms which may fall within the scope of the present invention, a preferred embodiment / embodiments of the present invention will now be described by way of example only with reference to the accompanying drawings, in which: [Brief description of the drawings]

[0072] [Figure 1] FIG. 1 shows an overview of the machine learning system and method for predicting plaque stability to provide clinical decision support software applications. [Diagram 2] FIG. 2 illustrates the training and testing process of the method of the present invention. [Diagram 3] FIG. 3 shows a graphical representation of the measurements of vessel tortuosity. [Figure 4]FIG. 4 shows the area of ​​endothelial shear stress in Pascals of a coronary artery segment below a particular threshold. [Diagram 5] FIG. 5 shows a graphical representation of the arc / angle measurements obtained from the OCT images. [Figure 6] FIG. 6 shows the area of ​​endothelial shear stress in coronary artery segments below a particular threshold in Pascals and the corresponding 18F-NaF TBR within the segment. [Figure 7] FIG. 7 shows a graph of the correlation between microcalcification training data and model output data according to the methods disclosed herein. [Figure 8] FIG. 8 shows the correlation between microcalcification test data and model output data according to the methods disclosed herein. [Figure 9] FIG. 9 shows the optimized fit to the maximum TBR using all the measurements in Table 1. [Figure 10] FIG. 10 shows the optimized fit to the maximum TBR using a subset of measurements (top); the average relative contribution of these five measurements to the model fit (bottom). [Figure 11] FIG. 11 shows the optimized fit to the maximum TBR using a subset of categorical measurements taken from intravascular OCT images (top); the average relative contribution of the OCT measurements to the model fit (bottom). [Figure 12] FIG. 12 shows an example of an overfitted model of a neural network that includes a two-layer feedforward network with sigmoid hidden neurons and linear output neurons. [Figure 13] FIG. 13 shows a block diagram illustrating an example computer system upon which an embodiment of system 100 may be implemented. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0073]

[0072] In the following description, it should be noted that similar or identical reference numbers in different embodiments indicate the same or similar features.

[0074]

[0073] The present invention is based on the discovery that any set of measurements can be fitted to a model via regression or machine learning techniques to obtain the exact microcalcification activity results typically measured using NaF PET. The systems and methods disclosed herein describe the unexpected realization of correlating sodium fluoride (NaF) uptake with features related to plaque anatomy, hemodynamic environment, etc., through the use of AI modeling methods from patient image data that utilize AI directly on the patient image data to identify and reconstruct anatomical structures, which are then used to extract derived data for use in a regression model to determine the microcalcification activity of the patient's arteries.

[0075]

[0074] Despite great clinical promise in determining microcalcification activity, obtaining measurements from current sodium fluoride-PET methods is costly, requires additional time for preparation and acquisition, complicates images, and is not widely available. Thus, sodium fluoride-PET is unlikely to reach routine clinical use. The systems and methods disclosed herein describe a framework for creating AI and regression models configured to provide the ability to determine information about microcalcification activity from data not related to sodium fluoride-PET in a way not previously considered or implemented. The systems and methods disclosed herein describe a solution to the technical difficulties of obtaining sodium fluoride-PET, addressing the long-standing problem of identifying those at highest risk of heart attack and those who would benefit most from intervention.

[0076]

[0075] Thus, the present invention is directed to a principle of general application in that it provides a method for measuring microcalcification activity in blood vessels (eg, coronary arteries) without the need to perform NaF PET imaging.

[0077]

[0076] The features of the present invention are described in more detail in the following sections of this description, which describe non-limiting aspects, embodiments and examples of the present invention. This description is included for the purpose of illustrating the present invention. The following description should not be understood as limiting the broad overview or disclosure of the present invention described above.

[0078] overview

[0077] Those skilled in the art will appreciate that the invention described herein is susceptible to variations and modifications other than those specifically described. The present invention includes all such variations and modifications. The present invention also includes, individually or collectively, all of the steps, features, compositions and compounds referred to or shown in this specification, and any and all combinations of the steps or features or any two or more of them.

[0079]

[0078] Each document, reference, patent application, or patent cited in this text is expressly incorporated herein by reference in its entirety, meaning that it should be read and considered by the reader as part of the text. Documents, references, patent applications, or patents cited in this text are not repeated in this document solely for the sake of brevity.

[0080]

[0079] Any manufacturer's instructions, descriptions, product specifications, and product sheets for the products mentioned in this specification or described in any document incorporated by reference into this specification are incorporated by reference into this specification and may be employed in the practice of this invention.

[0081]

[0080] The present invention is not limited in scope by any of the specific embodiments described herein. These embodiments are intended for illustrative purposes only. Functionally equivalent products, formulations and methods are clearly within the scope of the invention described herein.

[0082]

[0081] The invention described herein may include one or more ranges of values ​​(e.g., dosages, concentrations, etc.). A range of values ​​is understood to include all values ​​within the range, including the values ​​defining the range, and adjacent values ​​within the range that produce the same or substantially the same results as the values ​​immediately adjacent to the values ​​defining the boundaries of the range. Thus, unless otherwise indicated, the numerical parameters set forth in this specification and claims are approximations that may vary depending on the desired properties sought to be obtained by the present invention.

[0083]

[0082] It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the invention as claimed.

[0084]

[0083] In this application, the use of the singular also includes the plural unless specifically stated otherwise.

[0085]

[0084] The articles "a" and "an" are used herein to refer to one or to more than one (i.e., to at least one) of the object of the article. By way of example, "an element" refers to one element or to more than one element.

[0086]

[0085] In this application, unless otherwise stated, the use of "or" means "and / or." Furthermore, the use of the term "comprising" and other forms such as "comprises" and "includes" is not limiting. Also, unless otherwise stated, terms such as "element" or "component" include both elements and components that constitute a single unit and elements and components that constitute two or more subunits.

[0087]

[0086] As used herein and in the claims, the phrase "at least one" in reference to a list of one or more elements should be understood to mean at least one element selected from any one or more elements of the list of elements, but not necessarily including at least one and every element of each specifically listed in the list of elements, and not excluding any combination of elements of the list of elements. This definition also allows for the optional presence of elements other than those specifically identified in the list of elements to which the phrase "at least one" refers, whether related or unrelated to the specifically identified elements. Thus, as a non-limiting example, "at least one of A and B" (or, equivalently, "at least one of A or B" or, equivalently, "at least one of A and / or B") refers, in one embodiment, to including at least one A, optionally multiple As, where B is absent (and optionally including elements other than B); in another embodiment, to including at least one B, optionally multiple Bs, where A is absent (and optionally including elements other than A); in yet another embodiment, to including at least one A, optionally multiple As, and at least one B, optionally multiple Bs (and optionally including other elements), etc.

[0088]

[0087] The term "about" as used herein refers to an amount that varies by 30%, preferably 20%, more preferably 10% relative to a reference amount, and indicates and is within experimental error of the indicated value (e.g., within a 95% confidence interval of the mean) or within 10% of the indicated value (whichever is greater). The use of the word "about" to modify a numerical value merely indicates explicitly that the numerical value should not be interpreted as an exact value. When used to average all values ​​of a variable and the indicated value of the variable and to refer to a time interval representing weeks, "about 3 weeks" refers to 17 to 25 days, and about 2 to 4 weeks refers to 10 to 40 days.

[0089]

[0088] Throughout this specification, unless the context requires otherwise, the word "comprise", or variations such as "comprises" or "comprising", are understood to mean the inclusion of a specified integer or group of integers, but not the exclusion of any other integer or group of integers. It is also noted that in this disclosure and particularly in the claims and / or paragraphs, terms such as "comprise", "comprised", "comprising" and the like may have the meaning ascribed to them in U.S. Patent Law; for example, they may mean "include", "included", "including", etc.; and that terms such as "consisting essentially of" and "consisting essentially of" have the meaning ascribed to them in U.S. Patent Law, for example, allowing for elements not expressly recited, but excluding elements found in the prior art or that affect a basic or novel characteristic of the invention.

[0090]

[0089] For purposes of this specification, although the method steps are described in a sequence, that sequence does not necessarily imply that the steps are performed in chronological order in that sequence, unless there is another logical way to interpret the sequence.

[0091]

[0090] Furthermore, when features or aspects of the invention are described in terms of a Markush group, those skilled in the art will recognize that the invention also is described in terms of any individual members or subgroups of members of the Markush group.

[0092]

[0091] The terms "patient" and "subject" are used interchangeably and include mammals and non-mammals, such as primates, farm animals, pets, laboratory animals, captive wild animals, birds (including eggs), reptiles, and fish. Thus, the term refers to at least monkeys, humans, pigs, cows, sheep, goats, horses, mice, rats, guinea pigs, hamsters, rabbits, cats, dogs, chickens, turkeys, ducks, other poultry, frogs, and lizards.

[0093] The terms "treat" and "treatment" refer to prevention of a disorder, disease, or condition to which such term applies, or prevention or alleviation of one or more symptoms of such disorder or condition. This includes therapeutic treatments, prophylactic treatments, and applications in which a subject reduces the risk of developing a disorder or other risk factors. Treatment does not require a complete cure of the disorder, but includes embodiments in which symptoms, underlying risk factors, or the progression of the disorder are alleviated.

[0094]

[0093] Other definitions of selected terms used herein are described in the detailed description of the present invention and apply throughout. Unless otherwise defined, all other scientific and technical terms used herein have the same meaning as commonly understood by those skilled in the art to which this invention belongs.

[0095]

[0094] The various methods or processes outlined herein may be coded as software executable on one or more processors employing any one of a variety of operating systems or platforms. Moreover, such software may be written using any of a number of suitable programming languages ​​and / or programming or scripting tools, and compiled as executable machine code or intermediate code that runs on a framework or virtual machine.

[0096]

[0095] In this regard, the various inventive concepts may be embodied as a computer readable storage medium (or multiple computer readable storage media) (e.g., computer memory, one or more floppy disks, compact disks, optical disks, magnetic tapes, flash memories, circuitry in field programmable gate arrays or other semiconductor devices, or other non-transitory or tangible computer storage media) encoded with one or more programs that, when executed on one or more computers or other processors, perform methods for implementing the various embodiments of the invention discussed above. The one or more computer readable media may be portable such that the program or programs stored thereon can be loaded into one or more different computers or other processors to implement the various aspects of the invention discussed above.

[0097]

[0096] The terms "program" or "software" are used herein in a generic sense to refer to any type of computer code or set of computer-executable instructions that can be used to program a computer or other processor to implement various aspects of the embodiments discussed above. Furthermore, according to one aspect, it should be understood that one or more computer programs that, when executed, perform the methods of the present invention need not reside on a single computer or processor, but may be distributed in a modular fashion among a number of different computers or processors to implement various aspects of the present invention.

[0098]

[0097] Computer-executable instructions may take many forms, such as program modules, executed by one or more computers or other devices. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform particular tasks or implement particular abstract data types. Typically the functionality of the program modules may be combined or distributed as desired in various embodiments.

[0099]

[0098] The data structure may also be stored in a computer-readable medium in any suitable format. For ease of explanation, the data structure may be shown as having fields that are related through their location in the data structure. Such relationships may similarly be achieved by assigning to the storage of the fields locations in the computer-readable medium that convey the relationship between the fields. However, any suitable mechanism may be used to establish the relationship between information in the fields of the data structure, including the use of pointers, tags, or other mechanisms for establishing relationships between data elements.

[0100]

[0099] Also, various inventive concepts may be embodied as one or more methods, examples of which are provided. The operations performed as part of a method may be ordered in any suitable manner. Thus, while shown as sequential operations in the exemplary embodiments, embodiments may be constructed in which operations are performed in an order different from that shown, which may include performing some operations simultaneously.

[0101] Description of the embodiments

[0100] In the following description, it should be noted that similar or identical reference symbols in different embodiments indicate the same or similar features.

[0102]

[0101] Disclosed herein is a method for predicting vascular plaque stability, particularly coronary plaque stability, and an automated computer-implemented system and method for predicting plaque stability to provide a clinical decision support software application configured as a computer-implemented system 100 as outlined in Fig. 1, which uses a precision medicine patient-specific approach by analyzing intravascular images and biomechanical computer models to provide comprehensive data on plaque stability. This creates patient-specific data that allows clinicians to assess the risk of plaque rupture and tailor treatment plans to individualize medical care. Fig. 1 shows the overall framework, which will be explained more clearly below.

[0103]

[0102] The system 100 disclosed herein is specifically configured to enable rapid segmentation and annotation of intravascular patient image data (e.g., OCT) and link with computer models that calculate both shear and structural stresses to return data not achievable by other methods that have been shown to predict clinical events (Stone et al., 2016). In fact, Professor Peter Stone of Harvard Medical School, who has supported the use of biomechanical data in predicting coronary events, has stated that "characterization of plaque risk based solely on anatomy, while necessary, is not sufficient to predict high-risk plaques that may destabilize and cause new clinical events." "If plaque stress could be calculated in a routine and time-efficient manner in the catheter lab, this information could be very useful in identifying the highest-risk plaques and informing management decisions." The methods and computer-implemented systems disclosed herein provide a solution to this pressing clinical need.

[0104]

[0103] The system 100 disclosed herein is a user-friendly, semi-automated software tool for evaluating vascular and plaque features of OCT data in 2D and 3D that can create patient-specific 3D anatomical models from commercially available OCT imaging systems. Importantly, the system 100 can be used with any existing OCT imaging system.

[0105]

[0104] Intravascular OCT 103 uses near-infrared light to create images of the inside of the coronary arteries. This technology provides very high resolution images (pixel size 10-15 microns) and allows cardiologists to see the inside of the arteries in 10 times more detail than using the next best technology, intravascular ultrasound (IVUS; pixel size 100-150 microns), and up to 35 times more detail than state-of-the-art CT images (pixel size >350 microns). OCT also allows cardiologists to clearly see plaque inside the arteries, measure fatty and thrombus buildup, and make precise measurements before and after placing stents.

[0106] OCT Analysis: The system 100 is configured to enable the user to rapidly analyze plaque features and create pixel-perfect segmentation (i.e., marking) tools for two-dimensional analysis 111 of the lumen of the imaged artery, and to segment the artery, for example, through a machine learning (ML) lumen contour segmentation routine that automates the user workflow. Edge detection algorithms based on state-of-the-art machine learning tools (deep learning with capsules) automatically obtain data on the arterial contour and the size, shape, and location of the artery. Current tools are time-consuming and manual (e.g., slice by slice, over 1000 slices per patient). Thus, the system 100 can provide significant time savings compared to current plaque assessment tools, including automatic lumen contour segmentation. Observed typical performance of the system 100 shows that the mean arterial segmentation is comparable to current state-of-the-art machine learning models, but the processing of the images is significantly faster, and thus the system 100 produces segmentations indistinguishable from those obtained by a human operator, and is reproducible with real-time or near real-time processing.

[0107]

[0106] Current features of clinical interest include: lipid regions; surface calcium; deep calcium; plaque-free walls; thrombi; macrophages; microchannels; cholesterol crystals; and thin-capped fibrous atheroma. Each of these poses a risk of causing heart attack and potential death due to plaque rupture or erosion. Each of these biomarkers is available within the system 100 and the location of every data point is stored in 3D for use during 3D registration and mapping information between different 3D workspaces. Additional functionality may be included where relevant.

[0108]

[0107] Once the OCT image analysis is complete, the user can convert the 2D image data into a 3D image for further analysis 121. This converts the OCT data into a 3D reconstruction.

[0109]

[0108] Software Mapping: In most clinical scenarios, OCT image data is acquired in parallel with another modality such as CT 105 and / or angiogram 107. The system 100 is adapted to fuse the high-resolution OCT 103 with a low-resolution angiogram (vascular) 107 or computed tomography (CT) or CT angiogram (CTA) 105. The CT data 105 or angiogram 107 is easily converted 115 into a three-dimensional model of the arterial tree that provides data regarding arterial centerlines within the imaged arterial tree. Image data from different imaging modalities may be transformed into a common reference frame or coordinate system such that the combination of OCT data 103, CT / A data 105 and vascular data 107 provides a complete view of the imaged arterial tree, including the left ventricular muscle mass acquired from the CT / A data. Once the individual data sets are acquired, the system 100 is configured to co-register the data from the different studies into the same xyz workspace so that all the different modes of data can be qualitatively and quantitatively related to each other. Biomechanical simulations become possible when OCT data 103 is fused with vascular data 107 or CT data 105. The result is 3D geometry with exquisite detail in the OCT domain and novel 3D quantitative pathology data that is not obtainable with existing commercial software.

[0110]

[0109] If, for some reason, only one particular mode of data is available, the user will often be able to perform many of the functions of the system 100, but will be subject to some limitations in the analysis methods available; for example, if only OCT data 103 is available, the main limitation is that shear stress calculations are unreliable due to the lack of 3D centerline data from CT / A or vascular data.

[0111]

[0110] The software mapping process performed by the system 100 is configured to automatically interpolate the locations of the OCT image frames between anatomical landmarks. These landmarks (e.g., vascular branching points) may be acquired from other medical imaging modalities (e.g., CT / A or vascular data). Additionally, the CT data set may be utilized to acquire the left ventricular muscle mass, which is used to improve the boundary conditions of the CFD simulation.

[0112] Biomechanical Simulation - Structural Stress: Structural simulations can be performed at any location along the imaged vessel. These simulations are performed directly on the 2D OCT images, which are segmented by the analyst / user using semi-automated tools available in the system 100 to identify different regions of the plaque and vessel wall. Since it is not possible to know the material properties in vivo exactly, the system 100 uses a strategy based on tissue stiffness ratios, as in other regions of the cardiovascular system. This is a clinically applicable method and is unique to the system 100.

[0113] Microcalcification estimator: The presence of microcalcifications in coronary vessels is a predictor of future clinical events. This microcalcification activity is measured by the radioactive tracer on PET / CT. 18 F-sodium fluoride (NaF) uptake is used to image and quantitate the vascular endothelial cell endothelial cell endothelial cell. However, NaF-PET / CT imaging is expensive, not widely available, requires significant technical expertise to analyze the images, and also increases the radiation dose to the patient. Disclosed herein is a novel equation that predicts NaF uptake in vessel walls and plaques, and thus predicts microcalcification activity. Surprisingly, it has also been found that the equation significantly correlates with NaF uptake in vivo.

[0114]

[0113] The system and method disclosed herein enables customization of medical care by providing patients and their physicians with detailed OCT, biomechanical modeling, and predictive assessment of the likelihood of clinical events based on microcalcification activity in their arteries. This allows a departure from the current "one size fits all" approach used in hospitals, allowing better prevention methods to be tested for patients identified as high risk and de-escalating treatment for low risk patients. The system provides cardiologists with a new set of patient-specific tools, as it enables rapid qualitative and unparalleled quantitative offline analysis of OCT data, and provides biomechanical and microcalcification activity data not obtainable via other commercial means.

[0115] advantage

[0114] The computer-implemented system 100 disclosed in this specification is configured to provide distinct benefits and advantages over typical OCT image analysis tools integrated into OCT scanning equipment, and further configured to provide distinct advantages over third party available OCT software analysis tools, and to provide relevant measurements of vascular status from anatomical to functional status, including the following:

[0116]

[0115] Intraluminal shear stress (ESS), the biomechanical frictional force acting on the innermost layer of a blood vessel, is a known predictor of plaque growth, progression and clinical events. Several research tools exist for calculating ESS from vessels and CT-based 3D reconstructions (with or without the addition of OCT or IVUS), but most are cumbersome, require computational fluid dynamics expertise, and long computation times (e.g., 1-2 days on a typical workstation). System 100 utilizes a hybrid approach to calculate ESS that returns data within a clinically usable time frame.

[0117]

[0116] Plaque structural stress (PSS) is the force per unit area acting on a plaque. Plaque rupture occurs when PSS exceeds the plaque cap strength, and PSS also influences cellular activity that links to plaque remodeling, inflammation, erosion, cell proliferation and other activities related to plaque progression and stability. In vivo plaque rupture data shows that the location of maximum PSS coincides with the rupture site in over 80% of cases. Although data exists on the importance of PSS, PSS remains a research tool and has not been incorporated into commercially available software aimed at plaque analysis. This is in part due to a lack of knowledge of patient-specific material properties. System 100 circumvents this by using the principle of static determinism, which is widely utilized in other cardiovascular diseases such as aneurysms (Joldes et al., 2017) but not yet in coronary artery disease.

[0118]

[0117] Where OCT particularly excels is in plaque characterization. Its superior resolution over other diagnostic modalities means that features of the plaque wall can be identified and quantified down to the cellular level (e.g., the presence of macrophages). System 100 incorporates sophisticated tools to rapidly extract these features.

[0119]

[0118] Microcalcification activity in plaque walls has emerged as a powerful non-invasive indicator of future clinical events. This activity is detected as a radioactive tracer on positron emission tomography (PET) images. 18 F-sodium fluoride (NaF) uptake in arterial segments is measured. However, PET imaging is expensive, not easily accessible, difficult to interpret, and exposes patients to a lot of radiation. Disclosed herein is a novel method for predicting NaF uptake in arterial segments, and thus the likelihood of potentially future clinical events, without the need for PET imaging.

[0120]

[0119] Virtual stent placement is possible on the platform of the system 100. Since OCT provides unparalleled image resolution, stent planning is inherently more accurate. By determining the exact vessel dimensions, the exact stent selection is possible. Then, by selecting the appropriate 3D stent shape in the system 100, the stent can be virtually placed at the desired location in the vessel, after which flow simulation can be performed. This allows data on the performance of the stent to be obtained before surgery.

[0121] Vessel wall characterization is similar to plaque analysis, and the unparalleled image resolution means that the vessel wall can be rapidly identified and quantified. System 100 has developed an automated tool for lumen segmentation based on deep learning (artificial intelligence, AI) that currently outperforms the state of the art.

[0122]

[0121] Thin capsule measurement: Again, image resolution is a key factor. Thin capsules become clinically dangerous when they are less than 65 microns thick. Due to its resolution, OCT is the only diagnostic modality capable of measuring biomarkers of this risk.

[0123]

[0122] Multimodal imaging: The system 100 is configured to process any image data available to the clinician. CT and angiography images, even with vFFR, cannot provide accurate information about plague progression, erosion, and rupture. An ideal scenario would include a combination of imaging modalities (e.g., CCTA and OCT), but if the clinician wants to reduce the analysis performed using a single imaging modality (e.g., CCTA or angiography), this is possible with the system 100.

[0124]

[0123] Vascular bifurcations are included in the analysis of system 100. This provides accurate information about the flow within arterial segments and accounts for branching flow along the vessel.

[0125]

[0124] Fractional flow reserve (FFR) is the ratio of pressures upstream and downstream of a stenosis. If the pressure difference is greater than a certain threshold (e.g., 30%), intervention is considered. FFR using only image data is a current "hot topic" in cardiology, since it allows the measurement of FFR without the need for induction of hyperemic flow (forced increase in flow rate) or the presence of pressure measuring wires: two main factors that limit the uptake of standard FFR. Furthermore, image-based FFR has other major advantages. FFRCT (i.e., HeartFlow) is completely non-invasive, since it only requires a CT scan, but the calculation is quite time-consuming (times through HeartFlow are several hours). Most patients undergo angiography (invasive imaging) during routine clinical practice, but angiography-based FFR is much faster and cheaper (i.e., VIRTUheart and CAAS).

[0126]

[0125] FFR based on a combination of OCT and angiographic data (or only OCT if angiographic data is not available) is also possible.

[0127]

[0126] Rapid Timeframe: System 100 is designed to operate within a clinical timeframe and is intended to be a "push button". The methods used in system 100 have been verified to produce data generated using efficient simulation strategies (i.e., minutes of CPU time) that are comparable to data from longer simulations (i.e., days of CPU time).

[0128]

[0127] VR Visualization: The output from the system 100 is also VR compatible, providing an immersive view of the problem and the resulting data.

[0129] Aspects of the invention In a first aspect of the present invention, there is provided a method of measuring microcalcification activity in an artery, preferably a coronary artery, the method comprising: (a) measuring the steps of: (i) the presence and / or amount of coronary plaque or visible disease markers in the vascular tissue sample; (ii) the presence and / or amount of healthy tissue within said vascular tissue sample; (iii) one or more of the characteristics defining an abnormal hemodynamic environment within the blood vessel; (iv) one or more geometric features associated with vascular remodeling and affecting hemodynamics within the blood vessel; and / or (v) measuring one or more of the material properties that affect vascular hemodynamics; (b) calculating the microcalcification activity in the blood vessel as a function of the measurements obtained in step (a).

[0130] In one embodiment of the present invention, microcalcification activity in blood vessels, for example arteries such as coronary arteries or veins, is measured using Positron Emission Tomography (PET).

[0131]

[0130] According to the present invention, any set of measurements can be fitted to a model (via regression or machine learning techniques) to obtain accurate microcalcification activity results that are typically measured using NaF PET.

[0132] In one embodiment of the present invention, the presence and / or amount of vascular plaque is measured based on the measurement of well-established geometric disease markers from intravascular optical coherence tomography (OCT) images, specifically the presence of lipids, calcium and macrophages (bright spots) in the plaque. For example, measurements of the mean lipid arc [°], the mean calcium arc [°], and the mean bright spots are obtained. Additional geometric measurements indicative of disease relate to the vessel diameter, area, volume, arterial wall / layer thickness, tortuosity and eccentricity, as well as all combinations of these measurements. Similarly, these measurements can be obtained by any other imaging modality common in clinical practice.

[0133] In another embodiment of the present invention, the presence and / or amount of healthy tissue is measured based on the measurement of the amount of healthy arterial wall visible using intravascular OCT images: the plaque-free wall (PFW) is inversely correlated with disease. For example, a measurement of the mean arc [°] of the plaque-free wall is obtained. Similarly, this measurement can be obtained by any other imaging modality common in clinical practice.

[0134] In further embodiments of the present invention, measurements of abnormal hemodynamic environment such as blood-derived particle retention or abnormal WSS are typically estimated using computational fluid dynamics (CFD) simulations or other methods that can directly estimate wall shear stress (WSS) via imaging techniques such as MRI. For example, by measuring low shear area (LSA): area [%] of low WSS or high shear area (HSA): area [%] of high WSS or mean WSS [Pa]. Additional hemodynamic-derived metrics include, but are not limited to, oscillatory shear index (OSI), relative residence time (RRT), low oscillatory shear (LOS), endothelial activation potential (ECAP), velocity-derived field function (e.g., vorticity), pressure drop, or the gradient of any of the aforementioned metrics (e.g., gradient of WSS).

[0135] In yet another embodiment of the present invention, measurements of geometric features associated with vascular remodeling and influencing hemodynamics are preferably obtained from intravascular OCT images (perimeter and eccentricity) and coronary computed tomography angiography (CCTA), for example by measuring mean circumference [mm] (using OCT), mean eccentricity (using OCT), arterial wall / layer thickness, and / or ventricular muscle mass [g] (using CT).

[0136] In yet another embodiment of the present invention, material properties that affect hemodynamics, such as %hematocrit, can be measured during routine blood draws and used to calibrate the viscosity model used to calculate WSS in CFD. Other methods of measuring / estimating %hematocrit also apply here.

[0137]

[0136] In a preferred embodiment of the invention, the microcalcification activity measured by step (b) of the method allows for the assessment of treatment options based on one or more subject metrics.

[0138] In one embodiment of the invention, the imaging modalities used in obtaining the measurements of the method include computed tomography (CT), magnetic resonance imaging (MRI), ultrasound, intravenous ultrasound (IVUS), optical coherence tomography (OCT), single photon emission computed tomography (SPECT), PET, and NaF PET. Preferably, the measurements are obtained using NaF PET.

[0139] For a given set of measurements belonging to any of categories (i) to (v), obtained from a single or multiple sources (e.g., coronary artery imaging), a function may be obtained that estimates the microcalcification activity. Currently, an explicitly defined formula is used to estimate the microcalcification activity ( 18 The present study describes the contribution of any number of categorical measurements to the endothelial function (as measured using F-NaF PET).

[0140]

[0139] These formulas are described below in the example function for P (microcalcification activity (μCA) prediction), where formulas of the form P1 to P6 use categorical measurements (A1, ..., En) as arguments, and lower case letters describe coefficients determined using a fitting method (e.g., iterative optimization). Offset values ​​(constants) may be determined during fitting. Subscript n is the maximum coefficient or measurement index of the function arguments within each category (A-E). Note that with respect to previous / other descriptions / presentations of function arguments (i.e., general patient data) in this document, the function arguments here are grouped into categories. This is done to describe how the various functional forms aggregate the relevant measurements to minimize the numerical fit coefficient of the parametric function. The formulas can be implemented / fitted ignoring measurements and coefficients of certain categories if such data does not exist or is excluded. For simplicity, the formulas written here describe intermediate / iterative arguments / terms / categories as "...". TIFF2024529405000012.tif18170TIFF2024529405000013.tif37170TIFF2024529405000014.tif16170T IFF2024529405000015.tif16170TIFF2024529405000016.tif17169TIFF2024529405000017.tif33170 Format P 1 assumes that all measurements contribute independently to the function (unique coefficients). Format P 2 is multiplied by a unique exponent, looking at the measurements for a given category given the same proportional scaling factor. Format P 3 is of the form P by using the same exponent for measurements of a given category. 2 Simplify. Format P 4 and P 5 provide examples of categorically multiplicative forms, but these forms can increase measurement error and are sensitive to measurements with zero (or extreme) values. Format P 6illustrates the formats that the various groups of categories can use. Measurements from categories A and B are closely related and may share unique indices, but simulation results (category C) are considered exclusive.

[0141]

[0140] As explained below, other finite combinations of formulas may be used or generated to further optimize the method. However, all of these functions contain at most one exponential coefficient per measurement to simplify the fitting process. The formula (especially P 1 ) utilizes a power law formulation that has been used to describe allometric scaling relationships throughout biology, including the coronary blood supply. Simple power law relationships appear in fluid mechanics, where analytical solutions to the Navier-Stokes equations for flow inside pipes (which are physiologically relevant) result in quantities such as flow, pressure and friction / WSS depending on radius.

[0142] Optimized (Parametric) Fit: To determine the coefficients of a scalar equation such as P (above), the sum of squared errors may be minimized according to the Nelder-Mead Simplex Algorithm: a direct search method for multidimensional unconstrained minimization: TIFF2024529405000018.tif10170

[0143]

[0142] Here, TIFF2024529405000019.tif5170 is a vector of measured NaF uptake data, TIFF2024529405000020.tif4170 is a vector containing the estimates. This estimate is calculated using a scalar equation for P for each sample. In the first iteration of the algorithm, the coefficients of P are guessed (or set to any random value). After each iteration, the coefficients are updated until a minimum is found and the algorithm terminates. This occurs if the change in error is less than a specified tolerance. To aid the algorithm's ability to find the optimal set of coefficients, the input data (arguments for P) are normalized by their mean value.

[0144] Before providing specific examples, the main aspects to the methodology / framework are described more broadly below. Central to this is the training and testing of a predictive model, after which the model may be deemed suitable for use.

[0145]

[0144] Broadly, the process 200 includes the following steps as shown in FIG.

[0146] Step 1: Obtain training data 201 consisting of: [A] including a plurality of data relating to a plurality of patients, and a plurality of data which may further include image data and / or biomechanical data at one or more anatomical locations. Tr , B Tr , C Tr , D Tr ...] 203 general patient data, and Microcalcification activity data 205 (μCA) from multiple patients at one or more anatomical locations.

[0147] Step 2: Fitting a multivariate function / model 207, which includes calculating a function from the input patient data and microcalcification activity data that can estimate / predict μCA for new data acquired in the same way as the input training data, said function being: TIFF2024529405000021.tif6170

[0148] Step 3: Evaluate the multivariate model by evaluating the pre-fitted function f209 on the training set [A Te , B Te , C Te , D Te ... a set of estimates of microcalcification activity (μCA) for a new set of function inputs, a new set of common patient test data that was not included in Est ) step. TIFF2024529405000022.tif6170

[0149] Step 4: The error function E is used to calculate a set of estimates of microcalcification activity (μ Est ) to the known / corresponding value (μCA Te ) to calculate an error estimate 211: TIFF2024529405000023.tif6170

[0150]

[0149] Step 5: Check error estimates / assess model suitability step 213. If the error estimates meet a set of desired criteria, such as the required accuracy, precision and sensitivity to the input data, the model is considered fit for purpose and is used to predict microcalcification activity.

[0151]

[0150] If the error is small and the result is statistically significant, an estimated microcalcification activity is obtained. If the error is not small and / or the result is not statistically significant, further training is required by adding more training data.

[0152]

[0151] The following information provides background information regarding each aspect of the training and testing process listed in FIG.

[0153] Step 1(i) – General patient data 203 [A, B, C, D...]

[0152] Generic patient data may be derived from multiple patients and multiple data types, and may include image data and / or biomechanical data at one or more anatomical locations.

[0154]

[0153] Preferably, the general patient data includes patient data relevant to estimating microcalcification activity, the patient data including: 18 F-Sodium fluoride ( 18 This information includes information that may affect microcalcification activity detected / measured using F-NaF) positron emission tomography (PET).

[0155]

[0154] There is no restriction on the number of datasets collected for fitting or on the classification data within each dataset. For example, all measurements may or may not be geometric in nature. However, they must reasonably correspond to metrics / measurements expected to be associated with the atherosclerotic process and thus with microcalcification activity, and must be obtainable through image processing of medical imaging modalities using common image processing means used from clinical practice and / or, thus allowing the methodology to be readily applied.

[0156]

[0155] In a particular configuration, the geometric measurements correspond to image-based diameter measurements (standard measurement of vascular patency / health) in certain vessels prone to calcification processes, with extreme diameters being associated with unhealthy vasculature, and vessel size also being 18 It is expected that this will affect the surface area available for transporting the F-NaF tracer to the binding site. The geometric measurements are obtained by image processing of patient imaging data including one or more of computed tomography, optical coherence tomography, intravascular ultrasound, x-ray angiography, magnetic resonance imaging, or PET imaging.

[0157]

[0156] Apart from direct anatomical measurements of geometry, there are relevant measurements of vascular health or disease burden (e.g., coronary artery calcium score, etc.) and patient-specific measurements that have been demonstrated to support the progression of cardiovascular disease. These measurements include, for example, image-based measurements (i.e., computed tomography, optical coherence tomography, intravascular ultrasound, x-ray angiography, magnetic resonance imaging or PET imaging measurements) and non-image-based measurements such as patient history or blood sample data. Other relevant measurements are, for example, measurements of blood pressure, blood flow or local hemodynamic properties, and biomechanical measurements such as measurements of tissue stress. These types of metrics include: 18 It is expected to play a role in transporting the F-NaF tracer to its binding site and is widely associated with the progression of cardiovascular disease.

[0158]

[0157] Before being used in the model fitting process, the collected data may optionally undergo transformation / scaling to improve the performance of the fitting algorithm.

[0159] Step 1(ii) – Microcalcification Activity Data205 (μCA)

[0158] 18 F-NaF PET image data is acquired at one or more measurement scales from a PET scan of a patient's heart or other vascular region. This data should be collected for multiple patients and is the dependent variable in the fitting of a multivariate function / model (step 207 in FIG. 2).

[0160]

[0159] PET data was recorded according to standard methods for measuring 18 F-NaF PET data is recorded as an estimate of a standardized uptake value at each sample region / location. This value is preferably adjusted (normalized) for blood pool activity by measuring / assessing a standardized uptake value at a reference location. An example of this would be to take an average value from a region of interest in the right atrium. By doing this, the PET measurement process is standardized between patients and provides a measure of tissue-to-background ratio (TBR). In the coronary arteries:18 F-NaF PET is often reported as TBR or other similar uptake measurements, see, e.g., Coronary Microcalcification Activity (CMA) (Kwiecinski, J et al., J Am Coll Cardiol. 2020;75(24):3061-74).

[0161]

[0160] Furthermore, the measurement scale is 18 F-NaF PET refers to the manner in which data is sampled. The data may be obtained as a maximum value within a region of the patient's vasculature. Alternatively, data may be sampled at discrete length intervals / regions of interest along the patient's vessels. Examples of this could be sampling data every 5 cm along the centerline of the vessel, or sampling data between branching points, or sampling data every nth image, or sampling the maximum value of data per vessel or predefined anatomical segment / section. Data may be mapped as a continuous function by measuring with respect to a continuous variable such as spatial dimension (e.g., axial distance in a medical image stack or distance along the centerline of a coronary artery). Obtaining data in this manner allows for continuous functions to be evaluated in a specific manner (not predefined / during data collection) prior to the fitting step (e.g., obtaining the maximum / average value of the function in a particular region / interval). Additionally, the locality of the sampled data points (e.g., spatial dimension) may be considered an independent variable in the fitting of a multivariate function. This allows the fitted multivariate function to be: 18 This allows for the assessment of the spatial dependence of microcalcification activity measured using F-NaF PET, and in the process, general patient data also benefit from a similar spatial discretization when acquired from medical images.

[0162] To improve spatial / anatomical localization of measurements, PET images are co-registered with another image source (with secondary / clearer representation of patient-specific anatomy), such as contrast-enhanced computed tomography (to improve vascular appearance) to obtain a co-localization of each18 It is preferable to improve the registration of spatial data associated with F-NaF PET samples. To aid in this process, motion correction algorithms, such as elastic motion correction, may be used to better present the PET image data.

[0163]

[0162] The collected data may optionally undergo transformation / scaling before being used in the model fitting process to improve the performance of the fitting algorithm.

[0164] Step 2 - Fit a multivariate function / model 207

[0163] The fitting process 207 may be performed using any method for fitting a multivariate function, such as parametric or non-parametric regression, as will be appreciated by those skilled in the art. The examples herein include the Nelder-Mead Simplex method, but alternative optimization methods that expect the same or similar coefficients are available and suitable, as will be readily appreciated by those skilled in the art. However, in all cases, regardless of whether a machine learning method is utilized for the fitting process 207, training data and test data are required. Non-parametric regression is preferred because the form of the predictor equation (function f in FIG. 2) does not have a predefined form, but is determined / constructed from information derived from the data to be fitted. However, for this reason, more data is required than parametric regression. This requires that a portion of the training data 201 is saved as validation data 215 during the fitting process, as presented in FIG. 2. Although most of the training data 201 may be used to fit the model, this validation set (or subset) 215 is used to estimate the prediction error for model selection. The category of non-parametric regression includes methods such as kernel regression and machine learning support vector machines. Parametric fitting, on the other hand, uses a method where the form of the function is assumed / predetermined and the coefficients of the function are learned / determined. This may be performed using parametric machine learning algorithms or traditional optimization methods that find the minimum of an objective function (the "sum of squared errors"). In the case of non-linear functions, a particular example of this suitable for use in method 200 is a direct search method for multidimensional unconstrained minimization - such as the Nelder-Mead Simplex method. When performing parametric optimization (i.e., Equations 1-6), the form of the function being fitted may benefit from leveraging commonly used functional forms that have been used to describe relationships across biology, such as allometric scaling functions.

[0165] Step 3 - Evaluate the multivariate model209 As shown in FIG. 2, model evaluation 209 is important to test the accuracy and suitability of the model. The model is evaluated on a set of test data (general patient data: model inputs / arguments; A Te , B Te , C Te , D Te ...) This data should preferably be acquired from a broad set of patients at multiple sites and should be of sufficient size to ensure that the fitted model is not subject to simple errors such as sensitivity during extrapolation (non-physical values ​​obtained for data outside the scope of the training data). The test set 215 should not include the typical patient data 203 used during training 201. Additionally, for the same set of patients, microcalcification activity data 205 (μCATe) should also be acquired so that the errors of the fitted model can be quantified.

[0166] Step 4 - Calculate the error estimates211

[0165] Microcalcification activity (μCA) of the test data Est Once a set of estimates / predictions of microcalcification activity (μCA Te ) The difference between each of these sets of data gives the distribution of the errors. If the errors are normally distributed, then μCA Est and μCA Te A linear correlation with is a very simple way to evaluate model performance. The error distribution may simply be used to evaluate the accuracy (ideally centered around zero) and precision (ideally with low variance / range / spread) of the model. Other useful information may be generated from the error distribution, such as investigating the relationship between the error and the values ​​of the predictors or the fitted values: this helps to evaluate the sensitivity of the model to the inputs / outputs.

[0167] The output of the model may be converted into a discrete / nominal classification, providing another way in which the error estimates can be tested (e.g., sensitivity and specificity). In the case of binary classification, this is done through measuring true positives, false positives, true negatives and false negatives. A cut-off value is required to classify elevated microcalcification activity. This may be established relative to a set of control patients (not suspected of cardiovascular disease) and / or a tissue-to-background ratio threshold (e.g., a relative background value above 1).

[0168] Step 4(a) – Check the error estimates to assess the model’s fit.

[0167] If the error estimates meet a set of desired criteria, such as the required accuracy, precision and sensitivity to the input data, the model is considered fit for purpose and used to predict microcalcification activity.

[0169] Preferred Embodiments

[0168] In the following, an example of a specific embodiment of the present invention is presented to better understand the nature of the present invention. An example of a purely image-based approach using parametric model generation is the coronary vasculature. Here, typical patient data is obtained from intravascular optical coherence tomography (OCT) imaging data and coronary computed tomography angiography (CCTA) imaging data. Following a registration step with the (contrast-enhanced) CCTA imaging data (aligning both image spaces and objects within them), the microcalcification activity data is 18 F-NaF PET imaging is used to obtain the measurements of microcalcification activity. General patient data and measurements of microcalcification activity are sampled in different regions of the coronary vasculature: the main coronary segments described by commonly used coronary segment maps. Following the data collection phase, a model (parametric) is fitted and tested. All methods and results are detailed in Table A below.

[0170]

[0169] In this example, all training and test data are collected in the same phase.

[0171]

[0170] When the system is used to determine microcalcification activity from coronary CT angiography (CCTA) data, the system receives raw CCTA image data and then determines the quality of these image data by accessing associated image metadata: for example, slice thickness and pixel size of the CTCA acquisition must exceed certain thresholds. If the quality control checks are passed, the system classifies the CCTA data for use in the AI ​​training dataset. If applicable, the system compares morphology and plaque features identifiable in CCTA with corresponding features visible in invasive imaging (e.g., OCT) of the same patient. This is to verify that the features identifiable in CCTA are spatially correlated with the features identifiable in OCT. After quality control, data such as anatomical metrics and image-based pixel density distributions are extracted from the AI-derived geometry for use in the regression training model. If the quality control checks are not passed, the system returns an error indicating that the raw CCTA data cannot be used for training data or further purposes of the system.

[0172] Step 1(i) – General Patient Data 203 These data sets are considered independent variables (inputs or predictors) of the model and are labeled A, B, C, D, etc. Multiple measurements of each input are made. For three inputs (i.e., A, B, and C), each patient has multiple measurements for each of A, B, and C taken at multiple locations. For example, in the coronary vasculature, example inputs may include arterial tortuosity (Tort); low shear area (LSA), and plaque free wall (PFW). These are measured at each coronary vessel segment within the vasculature across the spatial region imaged by (and common to) all three imaging modalities.

[0173] Tortuosity (Tort): Tortuosity of the vessel lumen centerline 301 (FIG. 3) measured from CCTA. Diseased vessels tend to have more tortuosity. The vessel centerline is usually constructed from the wall boundary distance field or by calculating the wall distance using ray casting, and may be reliably constructed using many centerline algorithms. Prior to this process, the vessel boundary needs to be defined, here done by image segmentation: in this process, objects on the image are thresholded (masked) within the region of interest between Hounsfield units / pixel levels that contain the target object (vessel lumen) and do not contain other objects around. Once the centerline data is extracted, here centerline tortuosity is simply measured as the ratio of the total length LC 301 (i.e., the sum of the distances between successive points) along the centerline segment divided by the shortest distance LS 303 (straight line) between the centerline endpoints that bounds the region of interest where the patient-specific measurements are taken as shown in FIG. 3.

[0174] Low Shear Area (LSA): The percentage of vessel (segment) lumen surface area where the wall shear stress value is below a certain threshold (0.4 Pa is used in this example, but 1 Pa is also commonly used to represent low shear stress in the arterial system). This threshold is associated with stagnation of blood flow near the wall and increased monocyte wall adhesion. Low wall shear stress is associated with the initiation and progression of atherosclerosis. Low wall shear stress is calculated from the area defined by the vessel lumen boundary using computational fluid dynamics, similar to centerline reconstruction. For the current dataset, this can be performed using vessel boundaries segmented from CCTA or OCT images. Selecting the OCT image (due to its superior pixel resolution) allows the OCT vessel lumen boundary to be registered to the CCTA image space, the OCT vessel boundary to be given a curvature, and the OCT measurements to be spatially corresponded / aligned with the vessel segment defined by the coronary segment map from which each measurement area is defined. Below is an image showing the low shear regions (in Pascals) identified on the surface of the OCT-derived geometry after registration to CCTA image space (see Figure 4).

[0175]

[0174] Plaque Free Wall (PFW): Plaque Free Wall is an OCT measurement taken on an OCT image, e.g., image 500 in FIG. 5, as an angle 501 about lumen center 503 where the intima 505 and media 507 are healthy in areas where the arterial wall is clearly visible and unobstructed by plaque features that attenuate the OCT signal (which is inversely proportional to the presence of disease). The angle measurement method allows the PFW to be mapped to the vessel boundary such that it is expressed as a percentage of the lumen surface area of ​​the vessel segment. FIG. 5 is shown as an example of the PFW arc angle 501 superimposed on the OCT image 500.

[0176] Step 1(ii) – Microcalcification activity data205 The dependent variables (predicted variables / outputs) of the model were evaluated for each vessel segment in the common image space. 18 F-NaF PET images. Each segment also has a corresponding typical patient data measurement from each of the three categories mentioned above, Tort, LSA, and PFW.

[0177] FIG. 6 also displays LSA surface area data. 18 An example image 600 of a F-NaF PET segment measurement is shown: both data sets are acquired for the same anatomical region.

[0178]

[0177] Microcalcification activity is measured as the maximum tissue-to-background ratio (TBR) for each segment: the maximum standardized uptake value within the segment is normalized by the blood pool activity, where blood pool activity is measured as the mean standardized uptake value of the right atrium. The regions of interest for which TBR is measured on PET images include the coronary artery walls (as microcalcification activity occurs in the wall and not the lumen).

[0179] Step 2 — Fitting a Multivariate Function / Model Here, half of the collected data is used to train the model. The multivariate model is assumed to have a combination of the following power law equations, which are often used to represent allometric relationships: TIFF2024529405000024.tif5170 where a, b, c, d, e, and f are the coefficients of the model determined during model fitting, and μCA Est is the model estimate of microcalcification activity (maximum segment TBR value). In this example, to determine the coefficients, the sum of squared errors (ε) is minimized according to the Nelder-Mead simplex algorithm: TIFF2024529405000025.tif5170 where μCA Tr is a vector / array containing all measurements of the maximum segment TBR in the set of training data, and μCA Est is a vector / array of estimates for a given set of coefficients being optimized. Once the coefficients are generated, a correlation analysis is performed to evaluate how well the model fits the training data. Considering that the model was optimized to fit this data, this correlation is expected to be high. For first order polynomials, the fit results 701 are shown in Figure 7.

[0180] Step 3 – Evaluate the multivariate model, calculate error estimates, and assess model fit Once the coefficients of the model have been determined, they are evaluated on the remaining general patient data (data that was not included in the training data used to fit the model). In contrast to the training data, the first-order polynomial fit 801 of FIG. 8 to this data is slightly weaker but still significant (R 2 Tr=0.71;R 2 T e =0.70;p-value T e <0.0001).

[0181]

[0180] Furthermore, Ef = μCA Est -μCA Te The distribution of errors obtained from the μCA was normally distributed with a mean of 0.099 and a standard deviation of 0.24. TeGiven a mean value of 1.16, the model tended to overpredict microcalcification activity by ~9% with moderate accuracy. Furthermore, a negative association (Spearman's Rho) was found between PFW values ​​and error values ​​(Rho=-0.73, p-value<0.001), suggesting that the current model has a high tendency to underpredict microcalcification activity in healthier vessel segments when measured using the PFW metric.

[0182]

[0181] Given this information, the model may be useful for patients presenting with CCTA and OCT images of the coronary vasculature, but may benefit from excluding the PFW and / or using a different metric instead.

[0183]

[0182] Additional objects, advantages and novel features are described in the following description or will become apparent to those skilled in the art upon examination of the drawings and the following detailed description of certain non-limiting embodiments.

[0184] Working Example

[0183] In one example of the present invention, different category data is obtained according to Table 1. TIFF2024529405000026.tif110170

[0185] The following results are the maximum measured 18 FIG. 1 shows the linear regression between F-NaF PET uptake (maximum target-to-background ratio (TBR)) and estimated microcalcification activity P calculated using equation (1) for P1 above.

[0186] In Figure 9, all measurements from Table 1 are used in the fit and a comparison is performed at the coronary artery segment and coronary vessel (i.e., full OCT pullback) measurement scales. Note that the size of each segment used to obtain the measurements is different, and the maximum TBR at the vessel scale is the maximum TBR of all segments across the vessel.

[0187] In Figure 9, fits 901 and 903 show a strong linear relationship between the model and the data. At the vascular measurement scale, the number of degrees of freedom of the fitted model (23 coefficients) exceeds the number of data points (20) used to optimize / fit the model coefficients. Thus, the current example model almost certainly overfits the data: the resulting model may not be able to fit additional data or reliably predict future observations.

[0188] In Fig. 10, it can be seen that by reducing the number of categorical measurements used while ignoring geometrical measurements, the model performs well on the vascular measurement scale and with less risk of overfitting the data. However, performance on the segmental measurement scale is not maintained. In this group of variables, the normalized area (LSA [%)) of low WSS (<0.4 Pa) and the mean arc of the plaque-free wall (mean value, PFW Arc [°]) contribute most (on average) to the approximation of microcalcification activity.

[0189]

[0188] Furthermore, by using only OCT-derived metrics, a model can be generated that performs well at both measurement scales (Figure 11). Interestingly, when fitting at different measurement scales, the resulting models differ in the weights applied to the various geometric measures of circumference and eccentricity. The optimization process depends on the initial conditions and does not necessarily need to find a global minimum. However, model variations are not unexpected. Local geometric measures indicate segment location and vary throughout the coronary vasculature with the development of plaque phenotypes. Proximal vessel segments are usually larger than distal segments and high-risk plaques tend to form in proximal vasculature.

[0190]

[0189] The relationships of these geometric variables with TBR and LSA measurements in Table 2 show that, taken alone, both circumference and eccentricity are positively related to TBR and LSA. However, circumference shows a slightly stronger relationship, while eccentricity has a weaker correlation coefficient. It is clear that the strength of individual correlations does not reflect their contribution to the multivariate model, since the relationships between all variables affect the model results. TIFF2024529405000027.tif112170

[0191]

[0190] Table 2 shows the blood supply (LVM 3 / 4 Additional simple models for approximating LSA are presented that incorporate factors influencing α (α = 0.01) and viscosity (HCT). These models provide improved approximations of microcalcification activity TBR compared to geometric measurements alone, and are competitive with CFD measurements of LSA. Including these simple models in the previous six-parameter OCT-derived multivariate model (Figure 11) improved the correlation coefficient (R 2 Seg : 0.71~0.72, R 2 ves : 0.81 to 0.90), but with a power law scaling coefficient, an independent parameter (R 2 Seg =0.75;R 2 Ves = 0.88) for HCT and LVM 3 / 4 Just include the following:

[0192]

[0191] At this point, it can be concluded that a simple multivariate function such as P1 in equation (1) above can be tailored to provide a model that fits the data well while relying on a subset of easily obtainable categorical measurements, although other alternatives are available, including those in which the form of the multivariate function is not defined prior to fitting.

[0193]

[0192] This method is limited by the amount of consistent data available (here it is applied only to segment data). For the current data set, the method fits the training data well, similar to function P1 in Figures 9-11, but was unable to predict the test data.

[0194]

[0193] A particular alternative method is a machine learning method that performs data-driven fitting (non-parametric), and an equation for P is generated. A two-layer feed-forward network with sigmoid hidden neurons and linear output neurons is one implementation. If the data is consistent and there are enough neurons in the hidden layer of the model, this model can fit any multidimensional mapping problem well. This network is implemented using the Levenberg-Marquardt backpropagation algorithm and a default value of 10 neurons in the hidden layer. However, it is not immune to overfitting. The performance of the model was tested with various ratios of training, validation and test data. In some cases, the training data is perfectly fitted, but it is clear that the model is overfitted and does not perform well on future data, as shown in Figure 12, which shows an example of AI-based fitting.

[0195] This problem also occurs when fitting high-dimensional models of P1 (more than 3 arguments) to the training and test sets, so that the true predictive power of the models in Figures 10 and 11 is unknown without further data. However, some analyses were performed using lower-dimensional models (Tables 3 and 4 below). These models are promising but would benefit from further scrutiny of larger datasets. The LSA and PFW models show consistency across all training / test dataset ratios. All models perform well when a high proportion of data is provided for training. TIFF2024529405000028.tif91170TIFF2024529405000029.tif82170

[0196] The method presented is: 18 We show that measurable microcalcification activity detected using F-NaF PET can be predicted using measurements of the local hemodynamic environment, the presence or absence of coronary plaque, and related metrics. The absence of a plaque-free wall, a common marker of disease, and the presence of areas of low endothelial low shear stress were integral to the multivariate model discussed.

[0197] Implementation example - Hardware overview

[0196] According to one embodiment, the techniques described herein are implemented by at least one computing device. The techniques may be implemented in whole or in part using a combination of at least one server computer and / or other computing devices coupled using a network, such as a packet data network. The computing device may be hardwired to perform the techniques, or may include digital electronic devices such as at least one application specific integrated circuit (ASIC) or field programmable gate array (FPGA) that are permanently programmed to perform the techniques, or may include at least one general-purpose hardware processor that is programmed to perform the techniques according to program instructions in firmware, memory, other storage, or a combination thereof. Such a computing device may combine custom hardwired logic, ASICs, or FPGAs with custom programming to realize the described techniques. The computing device may be a server computer, a workstation, a personal computer, a portable computer system, a handheld device, a mobile computing device, a wearable device, a body-worn or embedded device, a smartphone, a smart appliance, an internetworking device, an autonomous or semi-autonomous device such as a robot or an unmanned ground vehicle or aerial vehicle, any other electronic device incorporating hardwired and / or program logic to implement the described techniques, one or more virtual computing machines or instances in a data center, and / or a network of server computers and / or personal computers.

[0198]

[0197] Figure 13 is a block diagram illustrating an example computer system on which one embodiment of the above-described system 100 may be implemented. In the example of Figure 13, a computer system 1300 and instructions for implementing the disclosed techniques in hardware, software, or a combination of hardware and software are represented diagrammatically, e.g., as boxes and circles, at the same level of detail commonly used by those skilled in the art to which this disclosure pertains to communicate about computer architectures and computer system implementations.

[0199]

[0198] Computer system 1300 includes an input / output (I / O) subsystem 1302, which may include a bus and / or other communication mechanisms for communicating information and / or instructions between components of computer system 1300 via electronic signal paths. I / O subsystem 1302 may include an I / O controller, a memory controller, and at least one I / O port. The electronic signal paths are represented in the drawings diagrammatically, for example, as lines, one-way arrows, or two-way arrows.

[0200] At least one or more hardware processors 1304 are coupled to the I / O subsystem 1302 for processing information and instructions. The hardware processors 1304 may include, for example, a general-purpose microprocessor or microcontroller and / or an embedded system or a dedicated microprocessor such as a graphics processing unit (GPU) or a digital signal processor or an ARM processor. The processors 1304 may comprise an integrated arithmetic logic unit (ALU) or may be coupled to a separate ALU. One or more of the hardware processors 1304 may be implemented as dedicated image processing means for segmenting, annotating and otherwise analyzing patient image data. Alternatively, the functionality of the image processing means may be shared between each of the hardware processors 1304.

[0201]

[0200] The computer system 1300 includes one or more units of memory 1306, such as a main memory, coupled to the I / O subsystem 1302 for electronically and digitally storing data and instructions executed by the processor 1304. The memory 1306 is also used to store patient image data and training data for retrieval by the processor 1304. The memory 1306 may include volatile memory, such as various forms of random access memory (RAM) or other dynamic storage devices. The memory 1306 may also be used to store temporary variables or other intermediate information during execution of instructions executed by the processor 1304. Such instructions, when stored in a non-transitory computer-readable storage medium accessible to the processor 1304, can render the computer system 1300 into a special-purpose machine customized to perform the operations specified in the instructions.

[0202]

[0201] The computer system 1300 further includes a non-volatile memory such as a read only memory (ROM) 1308 or other static storage device coupled to the I / O subsystem 1302 to store information and instructions for the processor 1304. The ROM 1308 may include various forms of programmable ROM (PROM), such as erasable PROM (EPROM) or electrically erasable PROM (EEPROM). A unit of persistent storage 1310 may include various forms of non-volatile RAM (NVRAM), such as FLASH memory, or solid-state storage, magnetic disks, or optical disks, such as CD-ROM or DVD-ROM, and may be coupled to the I / O subsystem 1302 to store information and instructions. The storage 1310 is an example of a non-transitory computer-readable medium that may be used to store instructions and data that, when executed by the processor 1304, cause the computer-implemented method to perform the techniques herein.

[0203]

[0202] The instructions in memory 1306, ROM 1308 or storage 1310 may include one or more sets of instructions organized as a module, method, object, function, routine or call. The instructions may be organized as one or more computer programs, operating system services, or application programs, including mobile apps. The instructions may include an operating system and / or system software; one or more libraries supporting multimedia, programming or other functionality; data protocol instructions or stacks for implementing TCP / IP, HTTP or other communication protocols; file formatting instructions for parsing or rendering files coded using HTML, XML, JPEG, MPEG or PNG; user interface instructions for rendering or interpreting commands in a graphical user interface (GUI), command line interface or text user interface; application software, such as an office suite, internet access applications, design and manufacturing applications, graphics applications, audio applications, software engineering applications, educational applications, games or other applications. The instructions may implement a web server, a web application server or a web client. The instructions may be organized as a presentation layer, an application layer, and a data storage layer, such as a relational database system, object store, graph database, flat file system, or other data storage, with or without Structured Query Language (SQL).

[0204]

[0203] The computer system 1300 may be coupled to at least one output device 1312 via the I / O subsystem 1302. In one embodiment, the output device 1312 is a digital computer display. Examples of displays that may be used in various embodiments include a touch screen display, or a light emitting diode (LED) display, or a liquid crystal display (LCD), or an e-paper display. The computer system 1300 may include other types of output device 1312 instead of or in addition to a display device. Examples of other output device 1312 include a printer, a ticket printer, a plotter, a projector, a sound or video card, a speaker, a buzzer or a piezoelectric device or other audible device, a lamp or LED or LCD indicator, a tactile device, an actuator or servo.

[0205] At least one input device 1314 is coupled to the I / O subsystem 1302 for communicating signals, data, command selections, or gestures to the processor 1304. Examples of input device 1314 include touch screens, microphones, still and video digital cameras, alphanumeric and other keys, keypads, keyboards, graphic tablets, image scanners, joysticks, clocks, switches, buttons, dials, slides, and / or various types of sensors such as force sensors, motion sensors, thermal sensors, accelerometers, gyroscopes, and inertial measurement unit (IMU) sensors, and / or various types of transceivers such as cellular or wireless such as Wi-Fi, radio frequency (RF), or infrared (IR) transceivers and global positioning system (GPS) transceivers.

[0206] Another type of input device is the control device 1316, which may perform other automatic control functions such as cursor control or navigation in a graphical interface on the display screen instead of or in addition to input functions. The control device 1316 may be a touchpad, mouse, trackball or cursor direction keys for communicating directional information and command selections to the processor 1304 and for controlling cursor movement on the display 1312. The input device may have at least two degrees of freedom in two axes, a first axis (e.g., x) and a second axis (e.g., y), that allow the device to specify a position in a plane. Another type of input device is a wired, wireless or optical control device, such as a joystick, wand, console, steering wheel, pedal, gear shift mechanism or other type of control device. The input device 1314 may include a combination of multiple different input devices, such as a video camera and a depth sensor.

[0207] In another embodiment, computer system 1300 may comprise an Internet of Things (IoT) device in which one or more of output device(s) 1312, input device(s) 1314, and controller(s) 1316 are omitted. Or, in such an embodiment, input device(s) 1314 may comprise one or more cameras, motion detectors, thermometers, microphones, earthquake detectors, other sensors or detectors, measurement devices or encoders, and output device(s) 1312 may comprise a dedicated display, such as a single-line LED or LCD display, one or more indicators, display panels, gauges, valves, solenoids, actuators, or servos.

[0208]

[0207] If computer system 1300 is a mobile computing device, input device 1314 may comprise a Global Positioning System (GPS) receiver coupled to a GPS module capable of triangulating multiple GPS satellites and determining and generating geographic location or location data, such as latitude and longitude values ​​for the geophysical location of computer system 1300. Output device 1312 may include hardware, software, firmware and interfaces for generating position report packets, notifications, pulse or heart rate signals, or other repetitive data transmissions identifying the location of computer system 1300, alone or in combination with other application specific data directed to host 1324 or server 1330.

[0209]

[0208] The computer system 1300 may implement the techniques described herein using customized hardwired logic, at least one ASIC or FPGA, firmware and / or program instructions or logic that, when loaded and used or executed in conjunction with the computer system, cause or program the computer system to operate as a special purpose machine. According to one embodiment, the techniques described herein are executed by the computer system 1300 in response to the processor 1304 executing at least one sequence of at least one instruction contained in the main memory 1306. Such instructions may be read into the main memory 1306 from another storage medium, such as the storage 1310. Execution of the sequence of instructions contained in the main memory 1306 causes the processor 1304 to perform the process steps described herein. In alternative embodiments, hardwired circuitry may be used in place of or in combination with software instructions.

[0210]

[0209] As used herein, the term "storage medium" refers to any non-transitory medium that stores data and / or instructions that cause a machine to operate in a specific manner. Such storage media may include non-volatile media and / or volatile media. Non-volatile media include, for example, optical or magnetic disks, such as storage 1310. Volatile media include dynamic memory, such as memory 1306. Common forms of storage media include, for example, hard disks, solid state drives, flash drives, magnetic data storage media, any optical or physical data storage media, memory chips, and the like.

[0211]

[0210] Storage media are distinct from transmission media, but may be used in combination with transmission media. Transmission media are involved in transferring information between storage media. For example, transmission media include coaxial cables, copper wire and optical fibers, including the wires that make up the bus of the I / O subsystem 1302. Transmission media may take the form of acoustic or light waves, such as those generated during radio wave and infrared data communications.

[0212]

[0211] Various forms of media may be involved in carrying at least one sequence of at least one instruction to the processor 1304 for execution. For example, the instructions may initially be carried on a magnetic disk or solid state drive of a remote computer. The remote computer can load the instructions into its dynamic memory and send the instructions over a communications link, such as a telephone line using fiber optic or coaxial cable or a modem. A modem or router local to the computer system 1300 can receive the data on the communications link and convert the data into a format readable by the computer system 1300. For example, a receiver, such as a radio frequency antenna or infrared detector, can receive the data carried in a radio or optical signal, and appropriate circuitry can provide the data to the I / O subsystem 1302, such as placing the data on a bus. The I / O subsystem 1302 carries the data to memory 1306, from which the processor 1304 retrieves and executes the instructions. The instructions received by the memory 1306 may optionally be stored in storage 1310 either before or after execution by the processor 1304.

[0213]

[0212] The computer system 1300 also includes a communication interface 1318 coupled to the bus 1302. The communication interface 1318 provides a two-way data communication coupling to a network link 1320 that is directly or indirectly connected to at least one communication network, such as a network 1322 or a public or private cloud on the Internet. For example, the communication interface 1318 may be an Ethernet networking interface, an Integrated Services Digital Network (ISDN) card, a cable modem, a satellite modem, or a modem that provides a data communication connection to a corresponding type of communication line, such as an Ethernet cable or any type of metal or fiber optic line or a telephone line. The network 1322 broadly represents a local area network (LAN), a wide area network (WAN), a campus network, an internetwork, or any combination thereof. The communication interface 1318 may comprise a LAN card that provides a data communication connection to a compatible LAN, or a cellular radiotelephone interface that is wired to transmit or receive cellular data according to a cellular radiotelephone wireless networking standard, or a satellite radio interface that is wired to transmit or receive digital data according to a satellite wireless networking standard. In any such implementation, communication interface 1318 sends and receives electrical, electromagnetic or optical signals via signal paths that carry digital data streams representing various types of information.

[0214]

[0213] The network link 1320 typically provides electrical, electromagnetic or optical data communication directly or through at least one network to other data devices, for example using satellite, cellular, Wi-Fi or BLUETOOTH technology. For example, the network link 1320 may provide a connection through a network 1322 to a host computer 1324.

[0215]

[0214] Furthermore, the network link 1320 may provide a connection through a network 1322 or to other computing devices through internetworking devices and / or computers operated by an Internet Service Provider (ISP) 1326. The ISP 1326 provides data communication services through a worldwide packet data communication network represented as the Internet 1328. A server computer 1330 may be coupled to the Internet 1328. The server 1330 broadly represents any computer, data center, virtual machine or virtual computing instance with or without a hypervisor, or a computer running a containerized program system such as DOCKER or KUBERNETES. The server 1330 may represent an electronic digital service implemented using multiple computers or instances and accessed and used by transmitting a web service request, a uniform resource locator (URL) string with parameters in an HTTP payload, an API call, an app service call, or other service call. Computer system 1300 and server 1330 may form elements of a distributed computing system including other computers, a processing cluster, a server farm, or an organization of other computers that cooperate to perform a task or execute an application or service. Server 1330 may comprise one or more sets of instructions organized as a module, method, object, function, routine, or call. The instructions may be organized as one or more computer programs, operating system services, or application programs, including mobile apps.The instructions may include operating system and / or system software; one or more libraries supporting multimedia, programming, or other functionality; data protocol instructions or stacks for implementing TCP / IP, HTTP, or other communications protocols; file format processing instructions for parsing or rendering files coded using HTML, XML, JPEG, MPEG, or PNG; user interface instructions for rendering or interpreting commands for a graphical user interface (GUI), a command line interface, or a text user interface; application software, such as an office suite, Internet access applications, design and manufacturing applications, graphics applications, audio applications, software engineering applications, educational applications, games, or other applications. Server 1330 may include web application servers hosting a presentation layer, an application layer, and a data storage layer, such as a relational database system with or without Structured Query Language (SQL), an object store, a graph database, a flat file system, or other data storage.

[0216]

[0215] Computer system 1300 can send messages and receive data and instructions, including program code, through the network(s), network link 1320 and communication interface 1318. In the Internet example, a server 1330 might transmit a requested code for an application program through the Internet 1328, ISP 1326, local network 1322 and communication interface 1318. The received code may be executed by processor 1304 as it is received, and / or stored in storage device 1310, or other non-volatile storage for later execution.

[0217]

[0216] The execution of instructions in the case described in this section may implement a process in the form of an instance of a computer program that is running and is composed of program code and its current activity. Depending on the operating system (OS), a process may be composed of multiple threads of execution that execute instructions simultaneously. In this regard, a computer program is a passive collection of instructions, but a process may be the actual execution of those instructions. Multiple processes may be associated with the same program; for example, opening multiple instances of the same program often means that multiple processes are running. Multitasking may be implemented to allow multiple processes to share the processor 1304. Although each processor 1304 or processor core executes a single task at a time, the computer system 1300 may be programmed to implement multitasking that allows each processor to switch between running tasks without waiting for each task to finish. In one embodiment, the switch may be performed if a task performs an I / O operation, if the task indicates that it is available to switch, or upon a hardware interrupt. Time sharing is implemented to allow fast response of interactive user applications by allowing multiple processes to appear to be running simultaneously using rapid context switches. In one embodiment, for security and reliability, the operating system may prevent direct communication between independent processes providing strictly mediated and controlled inter-process communication facilities.

[0218]

[0217] The term "cloud computing" is generally used herein to describe a computing model that enables on-demand access to a shared pool of computing resources, such as computer networks, servers, software applications and services, and enables rapid provisioning and release of resources with minimal administrative effort or service provider interaction.

[0219]

[0218] Cloud computing environments (sometimes referred to as cloud environments or clouds) may be implemented in a variety of ways to best suit different requirements. For example, in a public cloud environment, the underlying computing infrastructure is owned by an organization that makes its cloud services available to other organizations or to the general public. In contrast, a private cloud environment is typically intended for use only by or within a single organization. A community cloud is intended to be shared by multiple organizations within a community, while a hybrid cloud consists of two or more types of clouds (e.g., private, community, or public) tied together by data and application portability.

[0220]

[0219] In general, the cloud computing model allows some of the responsibilities previously provided by an organization's own information technology department to be provided instead as service layers within a cloud environment for use by consumers (either internal or external to the organization, depending on the public / private nature of the cloud). Depending on the specific implementation, the exact definition of the components or functions provided by or within each cloud service layer may vary, but common examples include: Software as a Service (SaaS), where the consumer uses software applications running on the cloud infrastructure, while the SaaS provider manages or controls the underlying cloud infrastructure and applications; Platform as a Service (PaaS), where the consumer can develop, deploy, and otherwise control their own applications using software programming languages ​​and development tools supported by the PaaS provider, while the PaaS provider manages or controls other aspects of the cloud environment (i.e., everything below the runtime execution environment). Infrastructure as a Service (laaS), where a consumer can deploy and run any software application and / or provision the processing, storage, network and other basic computing resources, while the laaS provider manages or controls the underlying physical cloud infrastructure (i.e., everything below the operating system layer).Database as a Service (DBaaS), where a consumer uses database servers or database management systems that run on the cloud infrastructure, while the DBaaS provider manages or controls the underlying cloud infrastructure, applications, and servers, including one or more database servers.

Claims

A method for measuring the microcalcification activity in a patient's vascular system, the method comprising: (a) a step of measuring patient data, comprising: (i) the presence and / or amount of coronary artery plaque or visible disease markers in a vascular tissue sample derived from the vascular system of the patient; (ii) the presence and / or amount of healthy tissue in the vascular tissue sample derived from the vascular system of the patient; (iii) one or more characteristics defining an abnormal hemodynamic environment within a blood vessel derived from the vascular system of the patient; (iv) one or more geometric characteristics associated with vascular remodeling and affecting the hemodynamics within a blood vessel derived from the vascular system of the patient, and / or (v) measuring patient data including one or more of one or more material properties affecting the vascular hemodynamics within a blood vessel derived from the vascular system of the patient; (b) predicting the NaF uptake within the vascular system of the patient as a function of the patient data. The method according to claim 1, wherein step (b) comprises calculating a NaF tissue-to-background ratio. The method according to claim 1, wherein the patient data is derived from patient image data. The patient image data is computed tomography data, optical coherence tomography data, intravascular ultrasound data, X-ray angiography data, magnetic resonance imaging data, angiography image data, or CT angiography image data The method according to claim 3, which is one or more of the above. The method according to claim 1, wherein the patient data includes biomechanical measurements. The biomechanical measurements are blood pressure, blood flow or local hemodynamic characteristics, and tissue stress, and the method according to claim 5, which is a biomechanical measurement selected from the group consisting of. Claim 7 The method according to claim 1, wherein the one or more geometric characteristics correspond to atherosclerotic processes and / or microcalcification activity. Claim 8 The method according to claim 1, wherein the one or more geometric characteristics correspond to image-based diameter measurements of vessels prone to calcification. A computer-implemented method for measuring the microcalcification activity in a patient's vascular system, the method comprising: (a) a step of measuring patient data, comprising: (i) the presence and / or amount of coronary artery plaque or visible disease markers in a vascular tissue sample derived from the vascular system of the patient; (ii) the presence and / or amount of healthy tissue in the vascular tissue sample derived from the vascular system of the patient; (iii) one or more characteristics defining an abnormal hemodynamic environment within a blood vessel derived from the patient's vasculature, (iv) one or more geometric characteristics associated with vascular remodeling and affecting the hemodynamics within a blood vessel derived from the patient's vasculature, and / or (v) measuring patient data including one or more of one or more material properties affecting the vascular hemodynamics within a blood vessel derived from the patient's vasculature; (b) predicting NaF uptake within the patient's vasculature as a function of the patient data using a trained machine learning model, regression model, or prediction model. A method comprising:

10. The method according to claim 9, wherein the trained machine learning model comprises a first trained regression model or prediction model.

11. The method according to claim 1, wherein the patient's vasculature is one or more of a coronary artery, carotid artery, cerebral artery, aorta, peripheral artery, or vein.

12. The method according to claim 1, wherein the patient data includes CT angiography image data.

13. A computer system, the computer system comprising: at least one processor; at least one memory device for storing patient data, the patient data comprising: (i) the presence and / or amount of coronary artery plaque or visible disease markers in a vascular tissue sample derived from the patient's vasculature, and / or (ii) the presence and / or amount of healthy tissue in the vascular tissue sample derived from the patient's vasculature, and / or (iii) one or more characteristics defining an abnormal hemodynamic environment within a blood vessel derived from the patient's vasculature, and / or (iv) one or more geometric characteristics associated with vascular remodeling and affecting the hemodynamics within a blood vessel derived from the patient's vasculature, and / or (v) with respect to one or more material properties affecting the vascular hemodynamics within a blood vessel derived from the patient's vasculature, the at least one processor is configured to calculate microcalcification activity in the blood vessel as a function of the patient data using a trained machine learning model, regression model, or prediction model; at least one memory device; and a prediction processor for accessing the AI-trained model of the patient data and predicting NaF uptake within the patient's vasculature. A computer system comprising:

14. The computer system according to claim 13, wherein the processor is adapted to calculate a NaF tissue-to-background ratio.

15. The computer system according to claim 13, wherein the patient data is derived from patient image data.

16. The patient image data is computed tomography data, optical coherence tomography data, intravascular ultrasound data, X-ray angiography data, magnetic resonance imaging data, angiography image data, or CT angiography image data of one or more of, the computer system according to claim 15.

17. The computer system according to claim 13, wherein the patient data includes biomechanical measurements.

18. The biomechanical measurements are blood pressure, blood flow or local hemodynamic characteristics, and biomechanical measurements selected from the group consisting of tissue stress, the computer system according to claim 17.

Images