Imaging diagnostic methods for cardiovascular disease

Abstract

An image processing method for quantitatively assessing myocardial blood flow and myocardial blood flow reserve, comprising: (a) image preprocessing steps including: (i) reconstructing dynamic cine 3D tomographic myocardial perfusion imaging (MPI) data; (ii) denoising; (iii) extracting a blood input function from a region of interest (ROI) of the left ventricular blood chamber; (iv) estimating the distribution volume (DV) obtained from the ratio of uptake rate to washout rate (K1 / k2) to improve estimation of K1 and total blood volume (TBV) and myocardial blood flow measurement; and (v) normalizing the data by dividing by the maximum value of the blood input function; (b) evaluating the preprocessed signals in (a) to generate K1 and TBV parametric maps using an artificial neural network; and (c) postprocessing the K1, k2, and TBV parametric maps, as well as resting and stress myocardial blood flow.

Description

Technical Field

[0001] The present invention relates to nuclear imaging and medicine for evaluating quantitative myocardial blood flow and / or myocardial flow reserve.

Background Art

[0002] Nuclear medicine uses radioactive substances for therapy and medical imaging. There are various types of diagnostic imaging that utilize a dose of radiopharmaceutical. A dose of radiopharmaceutical can be injected into a patient before or during a diagnostic imaging examination. The radiopharmaceutical is absorbed by or attached to cells of a target organ in the patient and emits radiation. A scanner or detector in the diagnostic imaging process detects the emitted radiation to generate an image of the organ. For example, to image body tissue such as the myocardium, Rb-82 (rubidium-82) is injected into the patient. The diagnostic imaging method detects the radiation of Rb-82, obtains clearer images of the myocardium, and diagnoses related problems.

[0003] Radioisotopes play an important role in the diagnosis and palliation of various medical conditions. Radioisotopes are used, for example, in the treatment of cancer 60 Co, in the treatment of hyperthyroidism 131 I, in breath tests 14 C, as a tracer in myocardial blood flow imaging 99m Tc and 82 Rb, among others.

[0004] Furthermore, rubidium-82 is generated in-situ by radioactive decay of strontium-82. A rubidium elution system utilizes a dose of rubidium-82 generated by elution in a radioisotope generator, and injects the radioactive solution into a patient.

[0005] In previous preclinical studies using dogs, it has been shown that the uptake of Rb-82 radionuclide by the myocardium is directly related to myocardial blood flow (MBF). Dynamic cine myocardial perfusion imaging (MPI) using radioisotopes can accurately predict myocardial blood flow (MBF) and myocardial flow reserve (MFR). Typically, estimation of MBF begins with segmentation of the left ventricular (LV) myocardium, followed by tracer kinetic modeling in a limited number of two-dimensional (2D) polar maps or segments. However, 2D polar map segments have drawbacks. Few 2D polar maps can correctly depict cardiac diseases associated with small regional flow impairments. Therefore, there is an alternative method for performing myocardial perfusion imaging by visualizing three-dimensional (3D) parametric maps of MBF. However, visualizing 3D parametric maps of MBF has several drawbacks. Furthermore, when identifying small regional flow impairments, they are not always visualized. In addition, generation of 3D parametric maps is time-consuming and lacks stability. These drawbacks make medical professionals hesitant to adopt 3D parametric maps of MBF for estimating myocardial blood flow. Therefore, there is a need for an alternative approach that can generate more stable 3D parametric maps in a shorter time for estimating myocardial blood flow and myocardial flow reserve.

[0006] Rb-82 PET imaging was performed using a single fixed dose for all patients, due to limitations of early-generation tracer delivery systems. The undesirable effects of such old PET imaging systems can be mitigated to some extent by using advanced, latest-generation Rb-82 elution systems. The present inventors have found that 3D parametric imaging of myocardial blood flow using Rb-82 PET can accurately estimate and / or predict myocardial blood flow (MBF) and / or myocardial flow reserve (MFR). Summary of the Invention

[0007] The present invention aims to provide an image processing method for quantitatively evaluating myocardial blood flow (MBF) and / or myocardial flow reserve (MFR).

[0008] The object of the present invention is to provide an alternative method for estimating and / or predicting quantitative myocardial flow (MBF) and / or myocardial flow reserve (MFR) using 3D voxel-wise parametric image data. This allows for better emphasis on small localized blood flow impairments. More precisely, the inventors of the present invention estimate myocardial flow (MBF) and / or myocardial flow reserve (MFR) and generate a voxel-wise parametric map by fitting the image series to a single tissue compartment model and comparing the projected data to a two-dimensional (2D) polar coordinate map of the left ventricle (LV). Thus, the present invention has the advantage of better emphasizing small localized blood flow impairments and generating localized blood flow and reserve values ​​that are independent of the segmentation of the LV polar coordinate map.

[0009] The object of the present invention is to provide an image processing method for quantitatively evaluating myocardial blood flow and / or myocardial blood flow reserve. An image reconstruction algorithm has been developed to improve image quality, and it is intended to enhance the quality of image reconstruction using an AI algorithm, thereby accelerating image processing during myocardial blood flow imaging (MPI), and thereby reducing nuclear medicine dose by up to 10 times.

[0010] The objective of this invention is to generate a blood flow parametric map in an AI model with high accuracy and within a timeframe acceptable for clinical use, thereby enabling future clinical implementation.

[0011] The object of the present invention is to provide an image processing method for evaluating quantitative myocardial blood flow and / or myocardial blood flow reserve, in which 3D parametric images of MBF generated by the present invention also recommend calcium scoring.

[0012] One embodiment of the present invention provides an image processing method for quantitatively evaluating myocardial blood flow and myocardial blood flow reserve, the image processing method comprising the following steps a to c. a. A step of preprocessing an image, (i) Reconstructing dynamic cine 3D tomographic myocardial blood flow imaging (MPI) data, (ii) Separate the values ​​of the voxel (i, j, k) at each time point ti (where i is 1 to N), (iii) Optionally, remove noise to improve image quality. (iv) Extracting a blood input function from a region of interest (ROI) in the left ventricle blood cavity or other arterial blood regions of interest. (v) To stabilize and improve the estimation of K1 and total blood volume (TBV) and subsequent myocardial blood flow measurement, estimate the volume of distribution (DV) obtained by the ratio of uptake rate to washout rate (K1 / k2), and, (vi) A step including normalizing the data by dividing by the maximum value of the blood input function. b. A step in which the individual signals preprocessed in step (a) are evaluated in order to generate K1 and TBV parametric maps using an artificial neural network. c. A step of post-processing K1 and TBV parametric maps, as well as resting and stressed myocardial blood flow, in order to estimate myocardial flow reserve (MFR) maps and / or coronary flow reserve (CFR) maps.

[0013] One embodiment of the present invention includes an image processing method for quantitatively evaluating myocardial blood flow and myocardial blood flow reserve, wherein the image reconstruction of the array is a dynamic cine series in which 3D tomographic voxels (i, j, k) obtained by PET reconstruction are arranged for a number of time steps ti (i is 1 to N).

[0014] Furthermore, one embodiment of the present invention includes an image processing method for quantitatively evaluating myocardial blood flow and myocardial blood flow reserve, wherein the input signal enters a multilayer perceptron (MLP), an artificial neural network and / or a convolutional neural network (CNN) and / or a long-term short-term memory (LSTM) network to simultaneously predict uptake rate (K1), washout rate (k2), and total blood volume (TBV).

[0015] Another embodiment of the present invention includes an imaging method for evaluating quantitative myocardial blood flow and / or myocardial blood flow reserve, wherein the imaging agent or radionuclide is administered by an automated generation and infusion system of radiopharmaceuticals produced by fission, neutron activation, cyclotron, generator and / or a combination thereof, and / or intravenous administration.

[0016] A further embodiment of the present invention includes an image processing method for quantitatively evaluating myocardial blood flow and myocardial blood flow reserve, the method comprising the following steps a to d. a. A step of preprocessing an image, (i) Reconstructing dynamic cine 3D tomographic myocardial blood flow (MPI) data, (ii) Separate the values ​​of the voxel (i, j, k) at each time point ti (where i is 1 to N), (iii) Optionally, remove noise to improve image quality. (iv) Extracting a blood input function from a region of interest (ROI) of the left ventricular blood space or another arterial blood region of interest. (v) To stabilize and improve the estimation of K1, k2, and total blood volume (TBV) and subsequent myocardial blood flow measurement, estimate the volume of distribution (DV) obtained by the ratio of uptake rate to washout rate (K1 / k2). (vi) A step including normalizing the data by dividing by the maximum value of the blood input function. b. A step in which time series data in voxels (i, j, k) and blood input functions are simultaneously applied to an artificial intelligence network to predict intake K1, k2, and TBV. c. A step of post-processing the K1, k2, and TBV parametric maps, the post-processing comprising (i) and (ii) below. (i) partial volume correction, and, (ii) Extracted fractions for estimating myocardial blood flow (MBF) at rest and under stress. d. A step of post-processing resting and stressed myocardial blood flow for estimating myocardial flow reserve (MFR) and / or coronary flow reserve (CFR). Here, the artificial neural network is selected from the group consisting of multilayer perceptrons (MLPs), artificial neural networks (ANNs), convolutional neural networks (CNNs), recurrent neural networks (RNNs), long-shorthand-short-term memory recurrent neural networks (LSTM-RNNs), gated recurrent unit (GRU) networks, generative adversarial networks (GANs), deep machine learning, and / or combinations thereof. [Brief explanation of the drawing]

[0017] Further features and advantages of the present invention will become apparent from the accompanying drawings and the following detailed description.

[0018] [Figure 1] This flowchart shows how the model acquires time series data in blood input functions and voxels, and outputs K1, k2, and TBV (total blood volume). [Figure 2] This flowchart shows how the model obtains blood input functions, time series data in voxels, and volume of distribution (DV), and outputs K1 and TBV (total blood volume). [Figure 3] This is a flowchart for model training. [Figure 4-19] This is a comparison of ground truth and a parametric map generated by AI. [Modes for carrying out the invention]

[0019] Currently, there is a need to improve radioisotope imaging procedures for estimating and / or predicting small localized dysfunction or impairments related to myocardial flow (MBF) and / or myocardial flow reserve (MFR), in which case the image series is fitted to a single tissue compartment model to generate a voxel-level parametric map comparing data projected onto a two-dimensional (2D) polar coordinate map of the target organ, such as the left ventricular (LV) myocardium. The inventors have found that generating localized flow and reserve values ​​can more clearly highlight small localized flow deficits and has the advantage of being independent of the segmentation of the LV polar coordinate map. The inventors have found that by using an alternative 3D parametric imaging method of myocardial flow using radioisotopes, myocardial flow (MBF) and / or myocardial flow reserve (MFR) can be accurately estimated and / or predicted. The present invention can be more readily understood from the following detailed description and embodiments of the invention.

[0020] As used herein, the term “imaging” refers to the techniques and processes used to create images of various parts of the human body for diagnostic and therapeutic purposes in digital health. These include radiography, radiofluoroscopy, magnetic resonance imaging (MRI), computed tomography (CT), medical ultrasound or endoscopic ultrasound elastography, tactile imaging, thermography, medical photography, and nuclear medicine functional imaging techniques (such as positron emission tomography (PET), dynamic positron emission tomography (SPECT), or single-photon emission computed tomography (SPECT)). Diagnostic imaging reveals the internal structure of the body and facilitates the diagnosis and treatment of diseases.

[0021] As used herein, the term “positron emission tomography (PET)” refers to a functional imaging technique that uses radioactive materials, known as radiotracers or radionuclides, to visualize and measure changes in metabolic processes and physiological activities, including blood flow, local chemical composition, and absorption. Commonly used radionuclide isotopes for PET imaging include Rb-82 (rubidium-82), WaterO-15 (oxygen-15), F-18 (fluorine-18), Ga-68 (gallium-68), Cu-61 (copper-61), C-11 (carbon-11), N-13 (ammonia-13), Co-55 (cobalt-55), Zr-89 (zirconium-89), Cu-62, Cu-64, I-124, Tc-99m (technetium), Tl-201 (thallium-201), and FDG (fluorodeoxyglucose). The preferred radionuclide is Rb-82, which has a half-life of 75 seconds.

[0022] As used herein, the term "SPECT" refers to single-photon emission computed tomography. Single-photon emission computed tomography is a nuclear medicine tomography technique that uses gamma rays to provide true 3D information. This information is typically displayed as cross-sectional slices of the patient, but can be freely reformatted or manipulated as needed. This technique requires the administration of a gamma-emitting radioisotope (nuclide) to the patient, usually by injection into the bloodstream. Marker radioisotopes typically bind to specific ligands to become radioligands, and due to their properties, they bind to specific types of tissues. This allows the ligand-radiopharmaceutical combination to be delivered to a site of interest in the body where it binds, and the ligand concentration is evaluated by a gamma camera. Examples of SPECT drugs include: 99m Tc Technetium-99m 99m Tc)-Sestamibi, 99m Examples include Tc-tetrophosmine, In-111, Ga-67, Ga-68, and Tl-201 (thallium-201).

[0023] As used herein, the term “diagnosis” refers to the process of identifying a disease, condition, or injury based on its signs or symptoms. Health history, physical examination, and tests such as blood tests, imaging, scans, and biopsies are used in diagnosis.

[0024] As used herein, the term “assessment” refers to a qualitative and / or quantitative assessment of blood perfusion in a part of the body or region of interest (ROI).

[0025] As used herein, the term “stress agent” refers to any agent used to induce stress in a patient or subject during an imaging procedure. Stress agents according to the present invention include vasodilators (e.g., adenosine, adenosine triphosphate and its mimics), A2A adenosine receptor agonists (e.g., regadenosone or the adenosine reuptake inhibitor dipyridamole), other pharmacological agents that increase blood flow to the heart, such as catecholamines (e.g., dobutamine, acetylcholine, papaverine, ergovin, etc.), or other external stimuli that increase blood flow to the heart, such as cold pressure, mental stress, or physical exercise.

[0026] As used herein, the terms “automated infusion system” or “radionuclide generation” and / or “infusion system” or “Rb-82 elution system” refer to a system for generating and / or injecting radionuclides or radiotracers for administration to a subject. An automated infusion system consists of a radioisotope generator, dose calibrator, computer, control unit, display unit, radioactivity detector, cabinet, cart, waste liquid bottle, sensors, shielding assembly, alarm or warning mechanism, tubing, source vial, diluent or eluent, and valves. An automated infusion system may be connected communicatively or electronically to an imaging system.

[0027] As used herein, the term "dose" refers to the dose of radionuclide required to perform imaging on a subject. The dose of the radionuclide administered to a subject is 0.01 MBq to 10,000 MBq.

[0028] As used herein, the terms "coronary artery disease" or "cardiovascular disease" refer to diseases of major blood vessels. Cholesterol-containing deposits (plaques) in coronary arteries and inflammation are the causes of coronary artery disease. Coronary arteries supply blood, oxygen and nutrients to the heart. Plaque accumulation narrows these arteries and reduces blood flow to the heart. Ultimately, reduced blood flow causes chest pain (angina pectoris), shortness of breath, and other signs and symptoms of coronary artery disease. Severe occlusion of an artery may lead to a heart attack. Myocardial infarction can be diagnosed by imaging diagnosis of myocardium and / or myocardial blood flow (MBF) under rest or pharmacological stress conditions, and regional myocardial blood flow can be assessed.

[0029] As used herein, the term "myocardial blood flow (MBF)" can be defined as the volume of blood passing through tissue at a certain rate. MFR is the ratio of MBF at maximal coronary vasodilation to MBF at rest, and is therefore affected by blood flow both at rest and during stress. MFR represents the relative reserve capacity of the coronary circulation.

[0030] As used herein, the terms "radionuclide" or "radioisotope" refer to an unstable form of a chemical element that emits radiation as it decays and becomes stable. Radionuclides may occur naturally or be produced in a laboratory. In medicine, they are used for imaging examinations and treatments.

[0031] As used herein, "Sr / Rb elution system" or " 82Sr / 82 The term "Rb elution system (82Sr / 82Rb elution system)" refers to an injection system that generates a solution containing Rb-82, measures the radioactivity in that solution, and injects that solution into a subject to conduct various studies on the subject's area of ​​interest.

[0032] As used herein, the term "linage counts" refers to the number of radioactive isotope decays acquired per unit time by a PET scanner.

[0033] As used herein, the terms "generator" or "radioisotope generator" refer to a hollow column within a radiation shielding container. The column is filled with an ion exchange resin, and radioisotopes are packed into the resin. The radionuclide generator according to the present invention is: 99 Mo / 99m Tc, 90 Sr / 90 Y, 82 Sr / 82 Rb, 188 W / 188 Re, 68 Ge / 68 Ga, 42 Ar / 42 K, 44 Ti / 44 Sc, 52 Fe / 52m Mn, 72 Se / 72 As, 83 Rb / 83m Kr, 103 Pd / 103m Rh, 109 Cd / 109m Ag, 113 Sn / 113m In, 118 Te / 118 Sb, 132 Te / 132 I, 137 Cs / 137m Ba, 140 Ba / 140 La, 134 Ce / 134 La, 144 Ce / 144 Pr, 140 Nd / 140 Pr, 166 Dy / 166 Ho, 167 Tm / 167m Er, 172 Hf / 172 Lu, 178 W / 178 Ta, 191 Os / 191m Ir, 194 Os / 194 Ir, 226 Ra / 222 Rn, 225 Ac / 213 Bi, and 64 Zn / 61 Selected from Cu.

[0034] As used herein, the term “eluant” refers to a liquid or fluid used to selectively elute daughter radioisotopes from a generator column.

[0035] As used herein, the term "eluate" refers to the radioactive eluate obtained after the daughter isotopes have been acquired from the generator column.

[0036] As used herein, the term “Controller” refers to a computer or part thereof programmed to perform specific calculations, execute instructions, and control various activities of the elution system, either based on user input or automatically.

[0037] As used herein, the term “activity detector” refers to a component used, for example, to determine the amount of radioactivity present in the eluate from a generator before administering the eluate to a patient.

[0038] As used herein, the term “Convolutional Neural Network (CNN)” refers to a system similar to a feedforward neural system. It is a type of artificial neural network used in time series and image processing, specifically designed to process pixel data. In deep learning, the Convolutional Neural Network (CNN / ConvNet) is a type of deep neural network most commonly applied to time series analysis, as well as natural and medical images. The blood input function is extracted manually or automatically from the region of interest (ROI). The average signal within the ROI is extracted at each time point to create a 1D signal blood input function.

[0039] As used herein, the term "multi-layer perceptron (MLP)" refers to an example of an artificial neural network widely used to solve a variety of problems, such as pattern recognition and interpolation.

[0040] As used herein, the term “recurrent neural network (RNN)” refers to a special type of artificial neural network adapted to function with time-series data or data containing sequences. Typical feedforward neural networks only handle data points that are independent of each other. RNN schemes include long-short-term memory (LSTM) or gated recurrent unit (GRU) networks. RNNs can work in conjunction with CNNs to form networks such as CNN-LSTM.

[0041] As used herein, the term “Gated Recurrent Unit (GRU)” refers to a type of Recurrent Neural Network (RNN) that uses less memory. It is a specific model of recurrent neural network that uses connections through a set of nodes to perform machine learning tasks related to memory and clustering. It features a gating mechanism for recurrent neural networks.

[0042] As used herein, the term "voxel" refers to a value on a regular grid in three-dimensional space in three-dimensional (3D) computer graphics. A voxel is short for volume pixel and is the smallest recognizable cubic portion of a 3D image. Voxelization is the process of adding depth to an image using a series of cross-sectional images called a volume dataset. These cross-sectional (or slice) images are composed of pixels. Pixels and slices together divide the image space into three dimensions (3D) of volume elements (voxels), forming a 3D scalar field.

[0043] As used herein, the term “Tissue Response Function (TRF)” refers to tracer motion modeling and is used to estimate physiological parameters such as myocardial blood flow (MBF) by mapping or transforming the shape of the “Arterial Input Function (AIF)” to the shape of the TRF.

[0044] As used herein, the term “blood input function” is commonly known as the arterial input function (AIF) and is defined as the concentration of a tracer in an artery measured over time with respect to a region of interest.

[0045] One embodiment of the present invention is an image processing method for quantitatively evaluating myocardial blood flow and myocardial blood flow reserve, comprising the following steps a to c. a. A step of preprocessing an image, (i) Reconstructing dynamic cine 3D tomographic myocardial blood flow imaging (MPI) data, (ii) Separate the values ​​of the voxel (i, j, k) at each time point ti (where i is 1 to N), (iii) Optionally, remove noise to improve image quality. (iv) Extracting a blood input function from a region of interest (ROI) of the left ventricular blood space or another arterial blood region of interest. (v) To stabilize and improve the estimation of K1, k2, and total blood volume (TBV) and subsequent myocardial blood flow measurement, estimate the volume of distribution (DV) obtained by the ratio of uptake rate to washout rate (K1 / k2), and, (vi) A step including normalizing the data by dividing by the maximum value of the blood input function. b. A step in which the individual signals preprocessed in step (a) are evaluated in order to generate K1 and TBV parametric maps using an artificial neural network. c. To estimate the myocardial flow reserve (MFR) map and / or coronary flow reserve (CFR) map, use K1, k2, ...K n , X 2 , R 2 Steps to post-process the volume of distribution (DV), TBV parametric maps, and myocardial blood flow at rest and under stress.

[0046] One embodiment of the present invention includes an image processing method for quantitatively evaluating myocardial blood flow and myocardial blood flow reserve, wherein the array image reconstruction is a dynamic series of 3D tomographic voxels (i,j,k) obtained by PET reconstruction arranged in a number of time steps ti (i is 1 to N).

[0047] Other embodiments of the present invention include an image processing method for quantitatively evaluating myocardial blood flow and myocardial blood flow reserve, wherein the region of interest (ROI) can be manually and / or automatically set.

[0048] Furthermore, one embodiment of the present invention includes an image processing method for quantitatively evaluating myocardial blood flow and myocardial blood flow reserve, which normalizes the data by dividing it by the maximum value of a blood input function.

[0049] One embodiment of the present invention includes an image processing method for quantitatively evaluating myocardial blood flow and myocardial blood flow reserve, wherein the data of the blood input function is normalized to a value between 0 and 1.

[0050] Another embodiment of the present invention includes an image processing method for quantitatively evaluating myocardial blood flow and myocardial blood flow reserve, wherein an input signal is fed into a multilayer perceptron and / or artificial neural network and / or generative adversarial network (GANs) and / or convolutional neural network (CNN) and / or long-short-term memory (LSTM) network to simultaneously predict uptake rate (K1), washout rate (k2), volume of distribution (DV), and total blood volume (TBV).

[0051] Furthermore, one embodiment of the present invention includes an image processing method for quantitatively evaluating myocardial blood flow and myocardial blood flow reserve, which generates local blood flow and reserve values ​​from images to highlight small local blood flow defects.

[0052] One embodiment of the present invention includes an image processing method for quantitatively evaluating myocardial blood flow and myocardial blood flow reserve, wherein the artificial neural network is selected from the group consisting of multilayer perceptrons (MLPs), artificial neural networks (ANNs), convolutional neural networks (CNNs), recurrent neural networks (RNNs), long-shorthand-short-term memory recurrent neural networks (LSTM-RNNs), gated recurrent unit (GRU) networks, deep machine learning, and / or combinations thereof.

[0053] Another embodiment of the present invention includes an image processing method for quantitatively evaluating myocardial blood flow and myocardial blood flow reserve, wherein an artificial neural network is input in multiple layers to estimate the volume of distribution (DV), and the multiple layers are selected from the group consisting of initial layers, intermediate layers, final layers of the network, or a combination thereof.

[0054] One embodiment of the present invention includes an image processing method for quantitatively evaluating myocardial blood flow and myocardial blood flow reserve, wherein the model predicts a k2 (washout rate) value.

[0055] Furthermore, one embodiment of the present invention includes an image processing method for quantitatively evaluating myocardial blood flow and myocardial blood flow reserve, wherein anisotropic diffusion filtering is performed using a Gaussian filter.

[0056] One embodiment of the present invention includes an image processing method for quantitatively evaluating myocardial blood flow and myocardial blood flow reserve, which estimates K1, k2, and total blood volume (TBV) by performing voxel-level operations using a 1D signal CNN-LSTM to generate a more accurate myocardial blood flow (MBF) estimate.

[0057] Another embodiment of the present invention includes an imaging method for quantitatively evaluating myocardial blood flow and myocardial blood flow reserve, wherein the images are characterized by the administration of Rb-82, O-15, N-13, F-18, Cu-62, Tc-99m, Tl-201, and / or combinations thereof.

[0058] One embodiment of the present invention includes an imaging method for quantitatively evaluating myocardial perfusion and myocardial perfusion reserve, wherein the images are characterized by administering Rb-82 to highlight small localized blood flow defects in resting and stressed PET perfusion imaging.

[0059] One embodiment of the present invention includes an imaging method for quantitatively evaluating myocardial blood flow and myocardial blood flow reserve, wherein a contrast agent or radionuclide is administered by an automated generation and infusion system of radiopharmaceuticals produced by fission, neutron activation, cyclotron and / or generator and / or intravenous administration.

[0060] One embodiment of the present invention includes an image processing method for quantitatively evaluating myocardial blood flow and myocardial blood flow reserve, and the automated radioisotope generation and injection system includes an Rb-82 elution system.

[0061] Other embodiments of the present invention include an image processing method for quantitatively evaluating myocardial blood flow and myocardial blood flow reserve, which predicts the ratio of myocardial blood flow under stress to myocardial blood flow at rest, fits the obtained images to a single-tissue compartment model or a multi-tissue compartment model to determine myocardial blood flow reserve and / or coronary artery blood flow reserve, and diagnoses disease status by performing the evaluation of the obtained images using multilayer perceptrons (MLPs), artificial neural networks (ANNs), convolutional neural networks (CNNs), recurrent neural networks (RNNs), long-shorthand-short-term memory recurrent neural networks (LSTM-RNNs), gated recurrent unit (GRU) networks, deep machine learning, deep neural networks, artificial neural networks, and / or combinations thereof.

[0062] Furthermore, one embodiment of the present invention includes an imaging method for quantitatively evaluating myocardial blood flow and myocardial blood flow reserve, the imaging method including positron emission tomography (PET), dynamic positron emission tomography, single-photon emission computed tomography (SPECT), magnetic resonance imaging (MRI), computed tomography (CT), and / or a combination thereof.

[0063] One embodiment of the present invention includes an image processing method for quantitatively evaluating myocardial blood flow and myocardial blood flow reserve, the method comprising the following steps a to c. a. A step of preprocessing an image, (i) Reconstructing dynamic cine 3D tomographic myocardial blood flow (MPI) data, (ii) Separate the values ​​of the voxel (i, j, k) at each time point ti (where i is 1 to N), (iii) Optionally, remove noise to improve image quality. (iv) Extracting a blood input function from a region of interest (ROI) of the left ventricular blood space or another region of interest of arterial blood. (v) To stabilize and improve the estimation of K1, k2, and total blood volume (TBV) and subsequent myocardial blood flow measurement, estimate the volume of distribution (DV) obtained by the ratio of uptake rate to washout rate (K1 / k2). (vi) A step including normalizing the data by dividing by the maximum value of the blood input function. b. A step in which time series data in voxels (i, j, k) and blood input functions are simultaneously applied to an artificial intelligence network to predict intake K1, k2, and TBV. c. A step of post-processing the K1, k2, and TBV parametric maps, the post-processing comprising (i) and (ii) below. (i) partial volume correction, and (ii) Extracted fractions for estimating myocardial blood flow (MBF) at rest and under stress. d. A step of post-processing resting and stressed myocardial blood flow data for estimating myocardial flow reserve (MFR) and / or coronary flow reserve (CFR).

[0064] A further embodiment of the present invention includes an image processing method for quantitatively evaluating myocardial blood flow and myocardial blood flow reserve, the method comprising the following steps a to d. a. A step of preprocessing an image, (i) Reconstructing dynamic cine 3D tomographic myocardial blood flow (MPI) data, (ii) Separate the values ​​of the voxel (i, j, k) at each time point ti (where i is 1 to N), (iii) Optionally, remove noise to improve image quality. (iv) Extract blood input functions from the region of interest (ROI) of left ventricular blood and other arterial blood regions of interest. (v) To stabilize and improve the estimation of K1, k2, and total blood volume (TBV) and subsequent myocardial blood flow measurement, estimate the volume of distribution (DV) obtained by the ratio of uptake rate to washout rate (K1 / k2). (vi) A step including normalizing the data by dividing by the maximum value of the blood input function. b. A step of simultaneously applying a time series of voxels (i, j, k) and a blood input function to an artificial intelligence network to predict intake K1, k2, and TBV. c. A step of post-processing the K1, k2, and TBV parametric maps, the post-processing comprising (i) and (ii) below. (i) partial volume correction, and (ii) Extracted fractions for estimating myocardial blood flow (MBF) at rest and under stress. d. A step of post-processing resting and stressed myocardial blood flow for estimating myocardial flow reserve (MFR) and / or coronary flow reserve (CFR). Here, the artificial neural network is selected from the group consisting of multilayer perceptrons (MLPs), artificial neural networks (ANNs), convolutional neural networks (CNNs), recurrent neural networks (RNNs), long-shorthand-short-term memory recurrent neural networks (LSTM-RNNs), gated recurrent unit (GRU) networks, generative adversarial networks (GANs), deep machine learning, and / or combinations thereof.

[0065] A further embodiment of the present invention includes an image processing method for quantitatively evaluating myocardial blood flow and myocardial blood flow reserve, the method comprising the following steps a to d. a. A step of preprocessing an image, (i) Reconstructing dynamic cine 3D tomographic myocardial blood flow (MPI) data, (ii) Separate the values ​​of the voxel (i, j, k) at each time point ti (where i is 1 to N), (iii) Optionally, remove noise to improve image quality. (iv) Extracting a blood input function from a region of interest (ROI) of the left ventricular blood space or another arterial blood region of interest. (v) To stabilize and improve the estimation of K1, k2, and total blood volume (TBV) and subsequent myocardial blood flow measurement, estimate the volume of distribution (DV) obtained by the ratio of uptake rate to washout rate (K1 / k2). (vi) A step including normalizing the data by dividing by the maximum value of the blood input function. b. A step in which the time series of voxels (i, j, k) and the blood input function are simultaneously applied to an artificial intelligence network to predict intake K1 and TBV. Here, the mean R 2 The value is between 0.9 and 1. c. A step of post-processing the K1 and TBV parametric maps, the post-processing comprising (i) and (ii) below. (i) partial volume correction, and (ii) Extracted fractions for estimating myocardial blood flow (MBF) at rest and under stress. d. A step of post-processing resting and stressed myocardial blood flow for estimating myocardial flow reserve (MFR) and / or coronary flow reserve (CFR). Here, the artificial neural network is selected from the group consisting of multilayer perceptrons (MLPs), artificial neural networks (ANNs), convolutional neural networks (CNNs), one-dimensional convolutional neural networks (1D-CNNs), recurrent neural networks (RNNs), long-shorthand-short-term memory recurrent neural networks (LSTM-RNNs), gated recurrent unit (GRU) networks, generative adversarial networks (GANs), deep machine learning, and / or combinations thereof.

[0066] Figure 1 is a flowchart showing how the model acquires time series data in the blood input function and voxels, and outputs K1, k2, and TBV (total blood volume).

[0067] Figure 2 is a flowchart showing how the model obtains the blood input function, time series in voxels, and volume of distribution (DV), and outputs K1 and TBV (total blood volume).

[0068] Figure 3 shows the model training flowchart, where the model takes a blood input function, a voxel time series, and distributed volume (DV) as inputs, and outputs K1 and TBV (total blood volume). K1, TBV, and DV are input into a single-tissue compartment model to generate a predicted voxel time series. The mean squared error between the predicted voxel time series and the input voxel time series is calculated and fed back into the model to update the model parameters.

[0069] Figure 4-19 This is a comparison of ground truth and an AI-generated parametric map. Figure 4-11 This is the result of the patient's 82Rb stress scan. . Non The ground truth K1 parametric map obtained using the linear least squares method is shown as K1 True. (See Figure 4) The K1 map predicted by the AI ​​model is shown as K1 Predicted (K1 pred). (See Figure 5) The difference between K1 True and K1 Pred is shown as K1 Difference (K1 diff). (See Figure 6) This shows the correlation between the overall images of K1 True and K1 Pred. (See Figure 7) . In Figure 7, Solid lines represent identity lines, while dotted lines represent fit lines of the correlation. Figure 12-19 This is the result of the patient's 82Rb resting scan. . Non Ground truth K1 parametric map (K1 True) using linear least squares method (See Figure 12) , K1 map predicted from AI model (K1 Pred) (See Figure 13) The difference between K1 True and K1 Pred (K1 Difference) (See Figure 14) Correlation between overall images of K1 True and K1 Pred (See Figure 15)This indicates that. In Figure 15, Solid lines represent identical lines, while dotted lines represent the fit of correlations.

[0070] One embodiment of the present invention includes an image processing method for quantitatively evaluating myocardial blood flow and myocardial blood flow reserve, in which consistent image quality is observed in the Rb-82 dose range of approximately 1 MBq to approximately 10,000 MBq.

[0071] One embodiment of the present invention includes an imaging method for evaluating quantitative myocardial blood flow and myocardial blood flow reserve, the imaging method including radiography, fluoroscopy, magnetic resonance imaging (MRI), computed tomography (CT), medical ultrasound or endoscopic ultrasound elastography, tactile imaging, thermography, medical photography, and nuclear medicine functional imaging techniques (e.g., positron emission tomography (PET), dynamic positron emission tomography, single-photon emission computed tomography (SPECT), etc.) and / or combinations thereof.

[0072] Another embodiment of the present invention includes an imaging method for quantitatively evaluating myocardial blood flow and myocardial blood flow reserve, wherein the dose of contrast agent to be administered is calculated by an automated generation and injection system.

[0073] A further embodiment of the present invention includes an image processing method for quantitatively evaluating myocardial blood flow and myocardial blood flow reserve, wherein the input signal is fed into a multilayer perceptron (MLP), artificial neural network (ANN), convolutional neural network and / or long-short-term memory (LSTM) network to simultaneously predict uptake rate (K1), washout rate (k2) and total blood volume (TBV), and the resulting image is evaluated to diagnose a disease state by deep neural network, artificial neural network, deep machine learning, or a combination thereof.

[0074] One embodiment of the present invention includes an imaging method for quantitatively evaluating myocardial blood flow and myocardial blood flow reserve, comprising the following steps a to d for performing and executing this test. a) A step of generating a sufficient amount of Rb-82 using an automated elution system of a Sr-82 / Rb-82 radionuclide generator, b) The step of administering the generated dose of Rb-82 to the patient, c) A step of performing an appropriate imaging procedure to obtain a higher quality image of a small area, and, d) A step of performing evaluation of acquired images and diagnosing a disease state using deep neural networks, artificial neural networks, deep machine learning, convolutional neural networks, recurrent neural networks, long-shorter memory-recurrent neural networks (LSTM-RNNs), generative adversarial networks (GANs), gated recurrent unit (GRU) networks, and / or combinations thereof.

[0075] One embodiment of the present invention includes an imaging method for quantitatively evaluating myocardial blood flow and myocardial blood flow reserve, the method further comprising administering a stress agent to a subject.

[0076] Another embodiment of the present invention includes an imaging method for quantitatively evaluating myocardial blood flow and myocardial blood flow reserve, wherein stress is induced by administering a stressor selected from adenosine, adenosine triphosphate, rigadenosone, dobutamine, dipyridamole, or exercise.

[0077] Furthermore, one embodiment of the present invention includes an image processing method for quantitatively evaluating myocardial blood flow and myocardial blood flow reserve, wherein the subject's weight is 1 to 300 kg, preferably 20 to 200 kg.

[0078] One embodiment of the present invention includes an imaging method for quantitatively evaluating myocardial blood flow and myocardial blood flow reserve, the automatic dose calculation further includes other parameters selected from the type of radioisotope, the half-life of the radioisotope, the lifetime of the generator (residual radioactivity in the radioisotope generator), the generator yield, the injection time, the flow rate, the time elapsed from the generation of the radioisotope to injection, the detector sensitivity of the scanning instrument, the resolution of the scanner, the type of camera or scanner, the acquisition time, the camera sensitivity, the type of disease to be diagnosed, the subject's condition or other special needs such as known allergies, cardiac function, liver function, kidney function, any other special needs, any other diseases of the subject, medications, the type of imaging technique used such as PET, SPECT, CT, MRI, and / or combinations thereof.

[0079] Other embodiments of the present invention include an image processing method for quantitatively evaluating myocardial blood flow and myocardial blood flow reserve, and the automated generation and injection system includes a cabinet, a radioisotope generator, a dose calibrator, a computer, a controller, a display device, an activity detector, a cabinet, a cart, a waste liquid bottle, sensors, a shielding assembly, an alarm or alarm mechanism, tubing, a source vial, a diluent or eluent, a valve, or a combination thereof. The automated generation and injection system generates radionuclides from a generator / column located within the system. By eluting the generator with a suitable eluent such as saline, a radionuclide eluate is produced from the generator and automatically administered by the system after radioactivity measurement. The dose is automatically calculated by the system based on input subject parameters. The system has the ability to calculate the flow rate and injection time according to the dose. The automated generation and injection system is 82 Sr / 82 It can be composed of any radionuclide generator suitable for administration to a subject, such as an Rb generator.

[0080] One embodiment of the present invention includes an imaging method for quantitatively evaluating myocardial blood flow and myocardial blood flow reserve, wherein an automated generation and injection system is electronically or communicatively coupled to the imaging system. The coupled imaging system can issue a warning if the image quality does not meet the standard and repeated administration or scanning is required.

[0081] Other embodiments of the present invention include an image processing method for quantitatively evaluating myocardial blood flow and myocardial blood flow reserve, and the automated generation and injection system is embodied in a portable (or mobile) cart housing some or all of the generator, processor, pump, memory, patient line, bypass line, positron detector, and / or calibrator, sensors, dose calibrator, activity detector, waste liquid bottle, controller, display, and computer. The cart, equipped with components for the generation and injection of radioisotopes, is mobile and can be moved to the patient's location or center, hospital, etc., as needed.

[0082] Furthermore, one embodiment of the present invention includes an imaging method for quantitatively evaluating myocardial blood flow and myocardial blood flow reserve, and the method for diagnosing / imaging hemoperfusion or blood flow in a region of interest includes: inputting subject parameters into a radioisotope generation and injection system; automatically calculating an appropriate dose; generating radionuclides from an automated generation or injection system based on the required dose to be administered; administering radionuclides to a subject in need; performing a PET or SPECT scan of the region of interest; automatically analyzing the images using computerized software; quantitatively evaluating blood flow in the region of interest; generating an automated report of the evaluation; and providing the subject with appropriate treatment options.

[0083] Other embodiments of the present invention include an imaging method for quantitatively evaluating myocardial blood flow and myocardial blood flow reserve, wherein the subjects are human subjects. The human subjects are male or female. The age of the subjects varies from 1 month to 120 years. The human subjects include neonates, children, adults and / or elderly populations.

[0084] Other embodiments of the present invention include imaging methods for quantitatively evaluating myocardial blood flow and myocardial blood flow reserve, and all numerical values ​​disclosed herein may vary by 1%, 2%, 5%, 10%, or up to 20%, when "about" is used. This variation may apply to all numerical values ​​disclosed herein.

[0085] Other embodiments of the present invention include an image processing method for quantitatively evaluating myocardial blood flow and myocardial blood flow reserve, wherein the 3D parametric images of MBF generated by the present invention have better image quality and pixel count. The present invention also provides medical staff with recommended alerts regarding the detection of coronary artery disease by analyzing the generated 3D parametric images of MBF and / or MFR.

[0086] Other embodiments of the present invention include imaging methods for quantitatively assessing myocardial blood flow and / or myocardial blood flow reserve, and calcium scoring by 3D parametric images of the MBF generated by the present invention is also recommended. The coronary calcium (CAC) score reflects the total area of ​​calcium deposition and calcium density. A score of 0 means that no calcium is found in the heart, which suggests a low likelihood of future heart attacks. If calcium is present, a higher score indicates a higher risk of heart disease. To evaluate the accuracy and reproducibility of the visual estimation of coronary calcium (CAC), positron emission tomography (PET), dynamic positron emission tomography (PET), hybrid positron emission tomography (PET), computed tomography (CT), and single-photon emission computed tomography (SPECT) / CT myocardial blood flow imaging (MPI) scans are performed.

[0087] Other embodiments of the present invention include an image processing method for quantitatively evaluating myocardial blood flow and / or myocardial blood flow reserve, wherein the image reconstruction algorithm has been developed to improve image quality, and is intended to improve the quality of image reconstruction using an AI algorithm, to speed up image processing, and to reduce the nuclear medicine dose in myocardial blood flow imaging (MPI) by up to one-tenth.

[0088] Other embodiments of the present invention include image processing methods for AI models for generating highly accurate blood flow parametric maps within a timeframe acceptable for clinical use, thereby enabling future clinical implementation. Each embodiment disclosed herein is applicable to each of the other disclosed embodiments. Accordingly, all combinations of the various elements described herein fall within the scope of the present invention.

[0089] The present invention will be better understood by referring to the experimental data described later, but those skilled in the art will readily understand that the specific experiments described in detail are merely illustrative examples of the present invention as fully described in the claims. [Examples]

[0090] Example 1 Patients were administered Rb-82. Scans were selected from 20 subjects / patients (N=20) with a wide range of uptake deficiency severity in Rb-82 stress PET perfusion imaging. Input signals included multilayer perceptrons (MLPs), artificial neural networks (ANNs), convolutional neural networks (CNNs), recurrent neural networks (RNNs), long-shorthand-short-term memory recurrent neural networks (LSTM-RNNs), gated recurrent unit (GRU) networks, generative adversarial networks (GANs), deep machine learning, and / or combinations thereof, to simultaneously predict uptake rate (K1), k2, and total blood volume (TBV). Then, 3D parametric images of K1, k2, and TBV were combined to estimate MBF and / or MFR.

[0091] Example 2 Rb-82 was administered to 40 patients (N=40) from two scanners (20 from GE Discovery 690 and 20 from GE Discovery 600) and identified from a 2019 cardiac PET study covering a wide range of defect severities in 82Rb stress PET. Discovery 690 data were split into training / validation / test sets in a 60:20:20 ratio. All Discovery 600 data constituted separate holdout test sets. This study used a 196×196×98mm area around the heart. 3 Image-derived arterial blood input function (AIF) and voxel time series / time activity curve (TAC) were used in the region. Motion modeling was performed using a one-tissue compartment model (1TCM) with a general nonlinear least squares (NLS) method to create a reference parametric map. The AIF and voxel TAC were input into a convolutional / long-short-term memory neural network (CNN-LSTM) to predict K1 and TBV, and the associated predicted TAC. The AI ​​model was optimized to minimize the mean squared error between the input TAC and the predicted TAC (Figure 3). The results are shown below.

[0092] result The AI ​​model accurately predicts K1 and TBV, and the average R 2 The values ​​were 0.998 and 0.991 for the Discovery 690, and 0.995 and 0.997 for the holdout test set of the Discovery 600. (Shown in Figure 4-19) When generating a parametric map using a typical central processing unit (CPU), the conventional NLS method took an average of 89.1 minutes, while the AI-enabled model described in Example 2 took only 7.21 seconds, a 741-fold improvement.

[0093] These two embodiments of the present invention relating to AI models are capable of generating highly accurate blood flow parametric maps within a clinically usable timeframe, and are therefore suitable for future clinical implementation.

Claims

1. Image processing method for quantitatively evaluating myocardial blood flow and / or myocardial blood flow reserve, a. A step of preprocessing an image, (i) Reconstructing dynamic cine 3D tomographic myocardial blood flow imaging (MPI) data, (ii) Separate the values ​​of voxels (i, j, k) at each time point ti (where i is 1 to N), (iii) As an option, remove noise to improve image quality. (iv) Extracting a blood input function from a region of interest (ROI) of the left ventricular blood space or another arterial blood region of interest (ROI), (v) K1, k ２ Furthermore, in order to stabilize and improve the estimation of total blood volume (TBV) and subsequent myocardial blood flow measurement, the ratio of intake rate to washout rate (K1 / k) is used. ２ To estimate the volume of distribution (DV) obtained by ), and (vi) Normalize the data of the blood input function by dividing it by the maximum value of the blood input function. Steps including, b. A step of evaluating the individual signals preprocessed in step (a) in order to generate parametric maps of K1 and TBV using an artificial neural network, and c. To estimate the myocardial flow reserve (MFR) map and / or coronary flow reserve (CFR) map, K1, k ２ and a parametric map of TBV, as well as a step of post-processing myocardial blood flow at rest and under stress, A method that includes this.

2. An image processing method for evaluating quantitative myocardial blood flow and / or myocardial blood flow reserve as described in claim 1, characterized by reconstructing the values ​​of dynamic cine 3D tomographic myocardial blood flow imaging (MPI) data arranged over each time point ti (i is 1 to N).

3. Image processing method for evaluating quantitative myocardial blood flow and / or myocardial blood flow reserve according to claim 1, wherein the region of interest (ROI) can be set by manual and / or automated procedures.

4. An image processing method for evaluating quantitative myocardial blood flow and / or myocardial blood flow reserve according to claim 1, comprising normalizing data of a blood input function by dividing it by the maximum value of the blood input function.

5. A method for image processing to evaluate quantitative myocardial blood flow and / or myocardial blood flow reserve according to claim 1, wherein the range of values ​​in the data normalization of the blood input function is from 0 to 1.

6. Image processing method for evaluating quantitative myocardial blood flow and / or myocardial blood flow reserve as described in claim 1, wherein the input signal is input to a multilayer perceptron (MLP) and / or artificial neural network (ANN) and / or convolutional neural network and / or long-term memory (LSTM) network, and the uptake rate (K1), k ２ A method for simultaneously predicting the total blood volume (TBV).

7. Image processing method for evaluating quantitative myocardial blood flow and / or myocardial blood flow reserve as described in claim 1, comprising generating local blood flow and reserve values ​​from an image to highlight small local blood flow defects.

8. A method for image processing to evaluate quantitative myocardial blood flow and / or myocardial blood flow reserve according to claim 1, wherein the artificial neural network is selected from the group consisting of multilayer perceptrons (MLPs), artificial neural networks (ANNs), convolutional neural networks (CNNs), recurrent neural networks (RNNs), long- and short-term memory recurrent neural networks (LSTM-RNNs), gated recurrent unit (GRU) networks, generative adversarial networks (GANs), deep machine learning, and / or combinations thereof.

9. Image processing method for evaluating quantitative myocardial blood flow and / or myocardial blood flow reserve according to claim 1, wherein an artificial neural network enters multiple layers to estimate the volume of distribution (DV), the multiple layers being selected from the group consisting of initial layers, intermediate layers, final layers of the network, or combinations thereof.

10. Image processing method for evaluating quantitative myocardial blood flow and / or myocardial blood flow reserve according to claim 9, wherein the model is k ２ A method for predicting the (extraction rate) value.

11. Image processing method for evaluating quantitative myocardial blood flow and / or myocardial blood flow reserve as described in claim 1, wherein the process is performed on a voxel-by-voxel basis using a 1D signal CNN-LSTM to process K1,k ２ A method for estimating total blood volume (TBV) and generating a more accurate estimate of myocardial blood flow (MBF).

12. Image processing method for evaluating quantitative myocardial blood flow and / or myocardial blood flow reserve according to claim 1, wherein the image is characterized by the administration of Rb-82, O-15, N-13, Cu-62-PTSM, 99m-Tc-Sestamibi, Tl-201, and / or combinations thereof.

13. Image processing method for evaluating quantitative myocardial blood flow and / or myocardial blood flow reserve according to claim 1, wherein the image is characterized by administering Rb-82 to highlight small localized blood flow defects in resting and stressed PET perfusion imaging.

14. A method for image processing to evaluate quantitative myocardial blood flow and / or myocardial blood flow reserve according to claim 1, wherein the contrast agent or radionuclide is administered by an automated generation and infusion system of radiopharmaceuticals produced by fission, neutron activation, cyclotron and / or generator and / or intravenous administration.

15. A method for image processing to evaluate quantitative myocardial blood flow and / or myocardial blood flow reserve according to claim 1, wherein the automated radioisotope generation and injection system comprises an Rb-82 elution system.

16. Image processing method for quantitatively evaluating myocardial blood flow and / or myocardial blood flow reserve as described in claim 1, comprising: predicting the ratio of myocardial blood flow under stress to myocardial blood flow at rest; fitting the obtained image to a single tissue compartment model to determine myocardial blood flow reserve and / or coronary artery blood flow reserve; and performing the evaluation of the obtained image using a multilayer perceptron (MLP), artificial neural network (ANN), convolutional neural network (CNN), recurrent neural network (RNN), long-shorthand-short-term memory recurrent neural network (LSTM-RNN), gated recurrent unit (GRU) network, deep machine learning and / or a combination thereof.

17. Image processing method for evaluating quantitative myocardial blood flow and / or myocardial blood flow reserve as described in claim 1, wherein the image processing method comprises positron emission tomography (PET), dynamic positron emission tomography, single-photon emission computed tomography (SPECT), magnetic resonance imaging (MRI), computed tomography (CT), and / or a combination thereof.

18. A myocardial imaging method for quantitatively evaluating myocardial blood flow and myocardial reserve, a. A step of preprocessing an image obtained by a rubidium-82 radioactive tracer, (i) Reconstructing dynamic cine 3D tomographic myocardial blood flow (MPI) data, (ii) Separate the values ​​of voxels (i, j, k) at each time point ti (where i is 1 to N), (iii) As an option, remove noise to improve image quality. (iv) Extracting a blood input function from a region of interest (ROI) of the left ventricular blood space or another arterial blood region of interest (ROI), (v) K1, k ２ Furthermore, in order to stabilize and improve the estimation of total blood volume (TBV) and subsequent myocardial blood flow measurements, the ratio of intake rate to washout rate (K1 / k) is used. ２ To estimate the volume of distribution (DV) obtained by ), and (vi) Normalize the data of the blood input function by dividing it by the maximum value of the blood input function. Steps including, b. Applying the time series at voxel (i, j, k) and the blood input function to an artificial intelligence network simultaneously to obtain the uptake K1, k ２ and predicting TBV. c. K1, k ２ and a step of post-processing the TBV parametric map, (i) Partial volume correction and, (ii) Extracted fractions for estimating myocardial blood flow (MBF) at rest and under stress, Steps including, d. A step of post-processing resting and stressed myocardial blood flow to estimate myocardial flow reserve (MFR) maps and / or coronary flow reserve (CFR) maps, comprising the step of analyzing MBF maps and / or CFR maps, A method that includes this.

19. A myocardial imaging method for quantitatively evaluating myocardial blood flow and myocardial reserve, a. A step of preprocessing an image, (i) Reconstructing dynamic cine 3D tomographic myocardial blood flow (MPI) data, (ii) Separate the values ​​of voxels (i, j, k) at each time point ti (where i is 1 to N), (iii) As an option, remove noise to improve image quality. (iv) Extracting a blood input function from a region of interest (ROI) of the left ventricular blood space or another arterial blood region of interest (ROI), (v) K1, k ２ Furthermore, in order to stabilize and improve the estimation of total blood volume (TBV) and subsequent myocardial blood flow measurements, the ratio of intake rate to washout rate (K1 / k) is used. ２ To estimate the volume of distribution (DV) obtained by ), and (vi) Normalize the data of the blood input function by dividing it by the maximum value of the blood input function. Steps including, b. Simultaneously apply the time series of voxels (i, j, k) and the blood input function to the artificial intelligence network, and obtain intake K1, k ２ and a step of predicting TBV, wherein the mean R ２ Steps where the value is between 0.9 and 1, c. A step of post-processing the K1 and TBV parametric maps, (i) Partial volume correction and, (ii) Extracted fractions for estimating myocardial blood flow (MBF) at rest and under stress, Steps including, d. A step of post-processing resting and stressed myocardial blood flow for estimating myocardial flow reserve (MFR) and / or coronary flow reserve (CFR), An artificial neural network is a step selected from the group consisting of multilayer perceptrons (MLPs), artificial neural networks (ANNs), convolutional neural networks (CNNs) and / or 1D convolutional neural networks (1D-CNNs), recurrent neural networks (RNNs), long-term memory recurrent neural networks (LSTM-RNNs), gated recurrent unit (GRU) networks, generative adversarial networks (GANs), deep machine learning, and / or combinations thereof. A method that includes this.

Images