Cardiac signal prediction method and system

Abstract

The present invention provides a cardiac signal prediction method and system. The cardiac signal prediction method comprises: collecting MRI data; predicting a cardiac cycle from the collected MRI data using an algorithm; and operating a portion of the collected MRI data corresponding to a selected portion of the predicted cardiac cycle.

Description

Technical Field

[0001] Aspects of the present disclosure relate generally to magnetic resonance imaging (MRI), and more particularly to predicting cardiac signals from MRI data. Background Art

[0002] MRI is a widely used medical technique that uses magnetic and radio frequency energy to produce images of regions of interest. During an MRI scan, a volume coil (e.g., a body coil) and a local coil (e.g., a surface coil) can acquire MR signals generated by nuclear relaxation inside the object being examined. Cardiac MRI can be used to produce detailed pictures of structures in and around the heart for use in evaluating the anatomy and function of the heart and detecting or monitoring heart disease. The output of the cardiac MRI process can be MRI data in the form of k-space data converted into a cardiac movie. In many applications of cardiac MRI, electrical signals generated by the heart are separately acquired and used as reference points for identifying cardiac phase information or timing in a cardiac movie.

[0003] For example, in cardiac MRI image reconstruction, cardiac phase information or timing can be used to trim the MRI image to a single cardiac cycle. Figure 1 A typical workflow 100 is shown, wherein MRI data is acquired 102 and a cardiac cine is reconstructed from the MRI data 104. A separately acquired ECG signal 106 is acquired and manually correlated with the image 104 to determine whether to remove or retain certain cardiac phases of the MRI data 108. In other post-processing applications, such as cardiac stress analysis, the shape and function of the myocardium may be of interest during the end-diastolic and / or end-systolic phases of the cardiac cycle. Figure 2 A typical workflow 200 is illustrated wherein MRI data in the form of previously acquired MRI images 202 may be manually labeled to identify cardiac phases, such as end-diastolic and / or end-systolic phases 204, and then used for analysis 206, such as cardiac stress analysis.

[0004] However, the correlation between cardiac signals and MRI data usually requires converting the MRI data into cardiac movies, which then requires the expertise of trained medical experts to manually correlate the cardiac activity detected by the MRI with specific types of cardiac signals. These manual activities can be time and labor intensive, and inconsistencies may occur. Summary of the invention

[0005] It would be advantageous to provide methods and systems that automatically correlate MRI data and cardiac signals consistently and without manual intervention.

[0006] According to one aspect of the present disclosure, a method includes acquiring MRI data; predicting a cardiac cycle from the acquired MRI data using an algorithm; and operating on a portion of the acquired MRI data corresponding to a selected portion of the predicted cardiac cycle.

[0007] The acquired MRI data may include k-space data.

[0008] The acquired MRI data may include image data.

[0009] The acquired MRI data may include undersampled MRI data.

[0010] The acquired MRI data may include ECG signals from the subject being studied acquired during the MRI scan.

[0011] The acquired MRI data may include video images of the subject being studied acquired during the MRI scan.

[0012] The acquired MRI data may include pulse data from the subject being studied acquired during the MRI scan.

[0013] The algorithm may include a deep learning model, which further includes a combination of CNN and RNN models, a GRU model, an LSTM model, a fully convolutional neural network model, a generative adversarial network, a back-propagation neural network model, a radial basis function neural network model, a deep belief network neural network model, and one or more of an Elman neural network model.

[0014] Operating on the portion of the acquired MRI data corresponding to the selected portion of the predicted cardiac cycle may include locating a data row in k-space of the acquired MRI data.

[0015] Operating on the portion of the acquired MRI data corresponding to the selected portion of the predicted cardiac cycle may include interpolating between rows of MRI data in k-space of the acquired MRI data.

[0016] Operating on the portion of the acquired MRI data corresponding to the selected portion of the predicted cardiac cycle may include interpolating between the MRI images.

[0017] Operating on the portion of the acquired MRI data corresponding to the selected portion of the predicted cardiac cycle may include performing a cardiac stress analysis using the portion of the acquired MRI data.

[0018] Operating on the portion of the acquired MRI data corresponding to the selected portion of the predicted cardiac cycle may include performing cine image reconstruction on the portion of the acquired MRI data.

[0019] The method may also include acquiring a cardiac signal corresponding to the MRI data; and using an algorithm to predict one or more predicted cardiac signals from the MRI acquisition data and the acquired cardiac signal.

[0020] The predicted portion of the cardiac cycle may represent any portion of the cardiac cycle.

[0021] The predicted portion of the cardiac cycle may represent one or more of the end-systolic or end-diastolic cardiac phases.

[0022] The predicted portion of the cardiac cycle may represent one or more of a P, Q, R, S, T, U, QRS complex, or PR interval cardiac wave.

[0023] According to one aspect of the present disclosure, a system includes receiving and control circuitry configured to operate an algorithm that predicts a cardiac cycle from MRI data and a processing engine configured to operate on a portion of the MRI data corresponding to a selected portion of the predicted cardiac cycle.

[0024] The acquired MRI data includes one or more of k-space data, image data or under-sampled MRI data.

[0025] The acquired MRI data includes one or more of ECG signals, video images or pulse data from the subject under study acquired during an MRI scan.

[0026] Among them, the algorithm includes a deep learning model, which further includes one or more of a combination of CNN and RNN models, a GRU model, an LSTM model, a fully convolutional neural network model, a generative adversarial network, a back-propagation neural network model, a radial basis function neural network model, a deep belief network neural network model, and an Elman neural network model.

[0027] Wherein, the processing engine operates on the portion of the acquired MRI data to locate data rows in k-space of the acquired MRI data.

[0028] Therein, the processing engine operates on the portion of the acquired MRI data to interpolate between one or more of the MRI images or MRI data rows in k-space of the acquired MRI data.

[0029] Wherein, the processing engine operates on the portion of the acquired MRI data to perform cine image reconstruction on the portion of the acquired MRI data.

[0030] Wherein, the deep learning model is further configured to predict the cardiac cycle based on a combination of the MRI data and a cardiac signal corresponding to the MRI data.

[0031] These and other aspects, implementations, and advantages of the exemplary embodiments will become apparent from the embodiments described herein considered in conjunction with the accompanying drawings. However, it should be understood that the description and drawings are designed for illustrative purposes only and not as a definition of the limitations of the disclosed invention, which should be referred to the appended claims. Additional aspects and advantages of the invention will be set forth in the following description and will be apparent in part from the description or may be understood by practice of the invention. Moreover, aspects and advantages of the invention may be realized and obtained by the means and combinations particularly pointed out in the appended claims. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] In the following detailed section of the present disclosure, the present invention will be described in more detail with reference to the exemplary embodiments shown in the accompanying drawings. These embodiments are non-limiting exemplary embodiments, wherein the same reference numerals represent similar structures in several views of the accompanying drawings, wherein:

[0033] Figure 1 A typical workflow is illustrated, in which MRI data is acquired and a cardiac movie is reconstructed from the MRI data;

[0034] Figure 2 A typical workflow is illustrated, wherein MRI data in the form of previously acquired MRI images may be manually labeled to identify cardiac phases and then used for analysis;

[0035] Figure 3 An exemplary process flow according to aspects of the disclosed embodiments is illustrated;

[0036] Figure 4 An exemplary MRI apparatus according to aspects of the disclosed embodiments is illustrated;

[0037] Figure 5 An exemplary MRI data source for implementing the disclosed embodiments is shown;

[0038] Figure 6 An exemplary deep learning model according to aspects of the disclosed embodiments is illustrated;

[0039] Figure 7 illustrates an exemplary architecture of a processing engine 306 in accordance with the disclosed embodiments; and

[0040] Figures 8 to 13 An exemplary process flow according to aspects of the disclosed embodiments is illustrated. DETAILED DESCRIPTION

[0041] In the following detailed description, many specific details are set forth by way of example in order to provide a thorough understanding of the relevant disclosure. However, it should be apparent to those skilled in the art that the present disclosure can be practiced without such details. In other cases, in order to avoid unnecessarily obscuring aspects of the present disclosure, known methods, processes, systems, components and / or circuits have been described at a relatively high level without detailed description. Various modifications to the disclosed embodiments will be apparent to those skilled in the art, and the general principles defined herein may be applied to other embodiments and applications without departing from the spirit and scope of the present disclosure. Thus, the present disclosure is not limited to the embodiments shown, but should be given the widest scope consistent with the claims.

[0042] It should be understood that the terms "system", "unit", "module" and / or "block" used herein are a method of distinguishing different components, elements, parts, portions or assemblies at different levels in ascending order. However, if the terms and other expressions can achieve the same purpose, the terms can be replaced by other expressions.

[0043] It should be understood that when a unit, module or block is referred to as being "on," "connected to," or "coupled to" another unit, module or block, it can be directly on, directly connected to, or coupled to another unit, module or block, or there may be intervening units, modules or blocks, unless the context clearly indicates otherwise. As used herein, the term "and / or" includes any and all combinations of one or more of the associated listed items.

[0044] Generally, the words "module", "unit" or "block" as used herein refer to logic specifically implemented in hardware or firmware, or to a collection of software instructions. The modules, units or blocks described herein may be implemented as software and / or hardware, and may be stored in any type of non-transient computer-readable medium or another storage device. In some embodiments, software modules / units / blocks may be compiled and linked into an executable program. It should be understood that software modules may be called from other modules / units / blocks or from themselves, and / or may be called in response to detected events or interruptions. Software modules / units / blocks configured to be executed on a computing device may be provided on a computer-readable medium, such as an optical disc, a digital video disc, a flash drive, a disk, or any other tangible medium, or may be provided as digital download material (and may be initially stored in a compressed or installable format that requires installation, decompression or decryption prior to execution). Such software code may be partially or completely stored on a storage device that executes a computing device so as to be executed by the computing device. Software instructions may be embedded in firmware, such as an erasable programmable read-only memory (EPROM). It should also be understood that hardware modules / units / blocks may be included in connected logical components, such as gates and triggers, and / or may be included in programmable units, such as programmable gate arrays or processors. The modules / units / blocks or computing device functions described herein may be implemented as software modules / units / blocks, but may be represented in hardware or firmware. Typically, the modules / units / blocks described herein refer to logical modules / units / blocks that may be combined with other modules / units / blocks or divided into submodules / subunits / subblocks, regardless of their physical organization or storage. This description may be applicable to a system, an engine, or a portion thereof.

[0045] The terms used herein are used only for the purpose of describing specific examples and embodiments and are not intended to be limiting. As used herein, the singular forms "a", "an", and "the" may also be intended to include the plural forms, unless the context clearly indicates otherwise. It should also be understood that the term "comprising" when used in the present disclosure specifies the presence of integers, devices, behaviors, described features, steps, elements, operations, and / or parts, but does not exclude the presence or addition of one or more other integers, devices, behaviors, features, steps, elements, operations, parts, and / or groups thereof.

[0046] These and other features and characteristics of the present disclosure, as well as the methods of operation and functions of the relevant elements of the structure and the economy of combination and manufacture of the parts, may become more apparent when considering the following description with reference to the accompanying drawings, all of which form a part of the present disclosure. However, it should be expressly understood that the drawings are for illustration and description purposes only and are not intended to limit the scope of the present disclosure. It should be understood that the drawings are not to scale.

[0047] The disclosed embodiments relate to a method comprising: acquiring MRI data; predicting a cardiac cycle from the acquired MRI data using an algorithm; and operating on a portion of the acquired MRI data corresponding to a selected portion of the predicted cardiac cycle.

[0048] The disclosed embodiments also relate to a system comprising: receiving and control circuitry configured to operate an algorithm to predict a cardiac cycle from MRI data; and a processing engine configured to operate on a portion of the MRI data corresponding to a selected portion of the predicted cardiac cycle.

[0049] See also Figure 3 , illustrates a schematic block diagram of an exemplary system 300 incorporating aspects of the disclosed embodiments. The system may include an MRI data source 302 for providing MRI data, such as one or more of cardiac k-space data, cardiac image data, a cardiac time series of images, ECG data, or other cyclic data. An algorithm 304 may generate a predicted portion of a cardiac signal from the MRI data, and a processing engine 306 may operate on a portion of the acquired MRI data corresponding to a selected portion of the predicted cardiac cycle portion. It should be understood that the components of the system 300 may be implemented in hardware, software, or a combination of hardware and software.

[0050] Figure 4 4 shows a schematic block diagram of an exemplary MRI device 302 for providing MRI data according to the disclosed embodiments. The MRI device 302 may include an MRI scanner 402, a control circuit 404, and a display 406. Figure 4As shown in the cross-sectional view in FIG. 4 , the MRI scanner 402 may include a magnetic field generator 408, a gradient magnetic field generator 410, and a radio frequency (RF) generator 412, all of which surround a table 414 on which the subject under study can be positioned. The MRI scanner 402 may also include: an ECG signal sensor 420, which is used to obtain MRI data in the form of ECG signals from the subject under study during the MRI scan; a camera 422, which is used to obtain MRI data in the form of video images of the subject under study during the MRI scan; and a pulse detector 424, which is used to obtain MRI data in the form of the subject's pulse during the MRI scan. In some embodiments, the MRI scanner 402 can scan a subject or a region of a subject. The subject can be, for example, a human body or other animal body. For example, the subject can be a patient. The region of the subject can include a portion of the subject. For example, the region of the subject can include tissue of the patient. The tissue may include, for example, lung, prostate, breast, colon, rectum, bladder, ovary, skin, liver, spine, bone, pancreas, cervix, lymph, thyroid, spleen, adrenal glands, salivary glands, sebaceous glands, testicles, thymus, penis, uterus, trachea, skeletal muscle, smooth muscle, heart, etc. In some embodiments, the scan may be a pre-scan for calibrating an imaging scan. In some embodiments, the scan may be an imaging scan for generating an image.

[0051] The main magnetic field generator 408 can generate a static magnetic field B0 and can include, for example, a permanent magnet, a superconducting electromagnet, a resistive electromagnet, or any magnetic field generating device suitable for generating a static magnetic field. The gradient magnetic field generator 410 can use a coil to generate a magnetic field in the same direction as B0 but with a gradient in one or more directions (e.g., along the X, Y, or Z axis in the coordinate system of the MRI scanner 402).

[0052] In some embodiments, the RF generator 412 can use an RF coil to transmit RF energy through a subject or a region of interest of the subject to induce an electrical signal in the region of interest. The resulting RF field is generally referred to as a B1 field, and is combined with a B0 field to generate an MR signal spatially located and encoded by a gradient magnetic field. The MRI scanner 402 may also include an RF detector 416 implemented using, for example, an RF coil, wherein the RF detector operates to sense the RF field and deliver the corresponding output to the receiving and control circuit 404. The function, size, type, geometry, location, number, or magnitude of the MRI scanner 402 may be determined or changed according to one or more specific conditions. For example, the MRI scanner 402 may be designed to surround a subject (or a region of a subject) to form a tunnel MRI scanner (referred to as a closed-bore MRI scanner) or an open MRI scanner (referred to as an open-bore MRI scanner).

[0053] ECG signal sensor 420 can be operated to obtain ECG signals from the subject under study during MRI scanning so that algorithm 304 can be used to subsequently identify the cardiac cycle and cardiac phase of the subject. Camera 422 can be operated to obtain video images of the subject under study during MRI scanning so that algorithm 304 can be used to subsequently identify the cardiac cycle and cardiac phase of the subject. During MRI scanning, the subject may be required to hold his breath and remain still in order to provide accurate MRI heart data while scanning. However, this may be difficult for many reasons, and the video image of the subject can be used as an input to algorithm 304 to further enhance the prediction of cardiac cycle and cardiac phase, especially to compensate for the subject's movement or breathing pattern during scanning that may adversely affect the collected MRI data. Pulse detector 424 can provide pulse data from the subject during MRI scanning, which can also be used as an input to algorithm 304 to further enhance the prediction of cardiac cycle and cardiac phase.

[0054] The receiving and control circuit 404 can control the overall operation of the MRI scanner 402, specifically, the overall operation of the magnetic field generator 408, the gradient magnetic field generator 410, the RF generator 412, and the RF detector 416. For example, the receiving and control circuit 404 can control the magnetic field gradient generator to generate a gradient field along one or more of the X, Y, and Z axes, and control the RF generator to generate an RF field. In some embodiments, the receiving and control circuit 404 can receive commands from, for example, a user or another system, and control the magnetic field generator 408, the gradient magnetic field generator 410, the RF generator 412, and the RF detector 416 accordingly. The receiving and control circuit 404 can be connected to the MRI scanner 402 via a network 418. The network 418 may include any suitable network that can facilitate information and / or data exchange of the MRI scanner 402. The network 418 may be and / or include a public network (e.g., the Internet), a private network (e.g., a local area network (LAN), a wide area network (WAN), etc.), a wired network (e.g., Ethernet), a wireless network (e.g., an 802.11 network, a Wi-Fi network, etc.), a cellular network (e.g., a long term evolution (LTE) network), a frame relay network, a virtual private network (“VPN”), a satellite network, a telephone network, a router, a hub, a switch, a server computer, and / or any combination thereof. Merely by way of example, the network 418 may include a cable network, a wired network, a fiber optic network, a telecommunications network, an intranet, a wireless local area network (WLAN), a metropolitan area network (MAN), a public switched telephone network (PSTN), a Bluetooth network, a ZigBee network, a near field communication (NFC) network, etc., or any combination thereof. In some embodiments, the network 418 may include one or more network access points. For example, the network 418 may include wired and / or wireless network access points, such as base stations and / or Internet exchange points, through which one or more components of the MRI scanner 402 may connect to the network 418 to exchange data and / or information.

[0055] According to some embodiments, the receive and control circuit 404 may operate an algorithm 304 for predicting one or more cardiac signals from MRI acquisition data and may include a processing engine 306 for operating on the MRI data. According to one or more embodiments, the processing engine 306 may be separate from the receive and control circuit.

[0056] Figure 5An exemplary MRI data source for implementing the disclosed embodiments is shown. The source of MRI data may include, but is not limited to, an MRI scanner 502, a storage 504 of MRI data (e.g., MRI slice images or other MRI device output), a storage 506 of k-space data from any number of MRI scans, an image storage 508, a storage 510 of ECG signals, a storage of video images of the subject under study acquired during an MRI scan, or one or more of any other sources of MRI data. In some embodiments, the MRI data may include any type of MRI image, any type of cyclic sequence data, for example, video data with individual images along a series of time points, k-space data with k-space along a series of time points, a combination of one or more of sequential ECG data at a series of time points, undersampled MRI data, and any other sequence or cyclic data from which a portion of a cardiac cycle can be predicted. In at least one embodiment, the data in the image storage 508 may include a Digital Imaging and Communications in Medicine (DICOM) image. The MRI data source may also include any number of local, remote, or cloud-based sources.

[0057] Figure 6 An example of an algorithm 304 used in the form of a deep learning model 600 is illustrated. The deep learning model 600 generally operates to predict portions of a cardiac cycle signal directly from MRI data. In this example, the deep learning model 600 may include a plurality of convolutional neural network 602 and recurrent neural network layers 6041-604 n MRI data points data t1, data t2 ... data tn from one or more MRI data sources 302 are provided to the convolutional neural network 602. As described above, the MRI data points data t1, data t2 ... data tn may include video data having individual images along each of a series of time points t1, t2 ... tn, k-space data having k-space along a series of time points, DICOM images, sequential ECG data at a series of time points, and any other sequence or cyclic data from which a portion of a cardiac cycle can be predicted, either individually or in any combination. The convolutional neural network 602 operates to pre-process the MRI data and provide the pre-processed sequence data to the recurrent neural network layers 6041-604 n . Multiple recurrent neural network layers 6041-604 n The pre-processed sequence data can be processed to generate predictions of the portion of the cardiac cycle for each time point t1, t2...tn. In some embodiments, the convolutional neural network 602 can be operated to extract features from the image or k-space data along a series of time points, and the recurrent neural network layers 6041 and 604 nThe extracted features can be used to predict the cardiac phase or ECG signal. The predicted portion of the cardiac cycle can represent any portion of the cardiac cycle. The predicted portion of the cardiac cycle can also represent one or more of the end-systolic or end-diastolic cardiac phases. In addition, the predicted portion of the cardiac cycle can represent one or more of the P, Q, R, S, T, U, QRS complex or PR interval cardiac phases.

[0058] The deep learning model 600 may be operable to predict the cardiac cycle in a prospective manner, or may be operable to predict the cardiac cycle in a retrospective manner.

[0059] When predicting in a prospective manner, the data collected at the current time point and before is used to predict the current cardiac phase of the entire cardiac cycle. During model training, the model is fed with input data (MRI data, ECG data, pulse data, etc.) and gold standard labels in the form of annotated phases of the cardiac cycle. No data is provided after the current time point. The model is trained to learn the relationship between the input data and the labels. During inference time, the trained model is used to predict the current phase in the cardiac cycle based on the data collected so far.

[0060] When predicting in a retrospective manner, input data is used to predict a cardiac phase in the middle of the cardiac cycle after all data has been collected (e.g., during a post-processing step). That is, the model can use data collected after the phase being predicted. The training and inference process is similar to the prospective case, except that data can be collected after the phase being predicted.

[0061] Regarding cardiac cycle prediction: The heart repeats this pattern between diastole and systole. These two phases are generally considered to be the landmarks that divide the cardiac cycle into phases. It should be understood that "predicting the cardiac cycle" refers to the ability of the model to predict any cardiac stage / phase (not limited to diastole and systole). The prediction can be in any format. For example, the model can be used to predict whether the current data represents the diastole phase. Continuous series of numbers (e.g., 0, 0.1, 0.2, etc.) can also be used in the prediction of the cardiac cycle or phase.

[0062] Although the deep learning model 600 is shown and described as including a combination of a convolutional neural network and a recurrent neural network, it should be understood that the deep learning model may include one or more gated recurrent units (GRUs), long short-term memory (LSTM) networks, fully convolutional neural network (FCN) models, generative adversarial networks (GANs), back-propagation (BP) neural network models, radial basis function (RBF) neural network models, deep belief network (DBN) neural network models, Elman neural network models, or any deep learning or machine learning model capable of performing the operations described herein.

[0063] Figure 7 An exemplary architecture of the processing engine 306 according to the disclosed embodiment is illustrated. The processing engine 306 may include computer readable program code stored on at least one computer readable medium 702 for performing and executing the processing steps described herein. The computer readable program code for performing the operations of the disclosed aspects may be written in any combination of one or more programming languages, including object-oriented programming languages ​​(such as Java, Scala, Smalltalk, Eiffel, JADE, Emerald, C++, C#, VB.NET, Python, etc.), conventional programming languages ​​(such as "C" programming language, Visual Basic, Fortran 2103, Perl, COBOL 2102, PHP, ABAP), dynamic programming languages ​​(such as Python, Ruby and Groovy), or other programming languages. The computer readable program code may be executed entirely on the processing engine 306, partially on the processing engine 306, as a stand-alone software package, partially on the processing engine 306 and partially on a remote computer or server, or entirely on a remote computer or server. In the latter scenario, the remote computer may be connected to the processing engine 306 via any type of network, including the networks mentioned above with respect to the network 418 .

[0064] The computer readable medium 702 can be a memory of the processing engine 306. In alternative aspects, the computer readable program code can be stored in a memory external to the processing engine 306 or remote from the processing engine 300. The memory can include magnetic media, semiconductor media, optical media, or any medium that is computer readable and executable. The processing engine 306 can also include a computer processor 704 for executing the computer readable program code stored on at least one computer readable medium 702. In at least one aspect, the processing engine 306 can include one or more input or output devices, generally referred to as user interfaces 706, which can be operated to allow input to the processing engine 306 or provide output from the processing engine 300, respectively. The processing engine 306 can be implemented in hardware, software, or a combination of hardware and software.

[0065] Figures 8 to 13 An exemplary process flow according to aspects of the disclosed embodiments is illustrated. Figures 8 to 12 In the processing flow, MRI data 802 may be collected from any suitable MRI data source, such as one or more of MRI data sources 502, 504, 506, 508, 510, and may optionally be collected as described below with reference to Fig.13ECG data 804 of the description. The acquired MRI data is provided to an algorithm 304, which operates to predict a cardiac cycle 806 from the acquired MRI data. The processing engine may then be used to operate on a portion of the acquired MRI data corresponding to a selected portion of the predicted cardiac cycle. In some embodiments, the acquired MRI data may be reduced before being provided to the algorithm. For example, certain irrelevant or unneeded cardiac phases or cardiac cycles may be removed, which may reduce the size of the data to be processed by the algorithm and reduce computational time when using the algorithm and computational time of the processing engine 306.

[0066] exist Figure 8 In an exemplary processing flow of , operations include: using the selected portion of the predicted cardiac cycle to locate MRI data lines 808 in k-space. During cardiac MRI, the heart beats and moves while being scanned, which makes it difficult to obtain a complete set of data points for the heart at any given time point. In practice, a complete cardiac cycle can be divided into several phases (e.g., 20 phases in one cardiac cycle). For each phase (e.g., the 10th phase), the MRI scanner 402 may only acquire a subset of the region of k-space because it is difficult to acquire all the data at once. For example, in some cases, only 1 / 3 of the entire k-space data can be acquired at a given time point. Using the cardiac cycle information, in the next cardiac cycle and the same phase (i.e., the 10th phase in the 2nd cardiac cycle), the MRI scanner 402 can acquire another subset of the region of k-space (e.g., another 1 / 3). Similarly, in the 10th phase of the 3rd cardiac cycle, the remaining portion of the k-space data can be acquired. Because the data typically repeats over the entire cardiac cycle, by using the predicted cardiac cycle and phase information, the initial and subsequently acquired k-space lines can be placed into appropriate corresponding positions.

[0067] exist Fig. 9In an exemplary processing flow of , the acquired MRI data includes undersampled MRI data, and the operation includes: interpolating 908 between MRI data rows in k-space using a selected portion of the predicted cardiac cycle. In a cardiac MRI scan, a user typically defines a plurality of cardiac phases into which each cardiac cycle can be divided, for example, a user can specify that each cardiac cycle is divided into 25 phases. However, since each subject has a different time interval for each cardiac cycle, and an MRI scanner typically acquires k-space data at a set consistent rate that is different from the subject's cardiac cycle, the MRI scanner can only generate k-space data for a subset of phases. In an example where each cardiac cycle is divided into 25 phases, the MRI scanner can only generate k-space data for 20 phases. As a result, data is interpolated between the phases to produce a specified number of phases, such as interpolating 20 phases into 25 phases. In order to perform the interpolation, the predicted cardiac cycle information for each phase and the predicted time intervals between each phase are required.

[0068] exist Fig.10 In an exemplary processing flow of , the acquired MRI data includes undersampled MRI data, and the operation includes: interpolating between MRI images using a selected portion of the predicted cardiac cycle 1008. As described above, the user typically defines a plurality of cardiac phases into which each cardiac cycle can be divided, for example, the user can specify that each cardiac cycle is divided into 25 phases. However, since each subject has a different time interval for each cardiac cycle, and the MRI scanner typically generates image data at a set consistent rate that is different from the subject's cardiac cycle, the MRI scanner can only generate image data for a subset of phases. In an example where each cardiac cycle is divided into 25 phases, the MRI scanner can only generate image data for 20 phases. As a result, data can be interpolated between phases to produce a specified number of phases, such as interpolating 20 phases into 25 phases. In order to perform interpolation, predicted cardiac cycle information for each phase and predicted time intervals between each phase are required.

[0069] exist Fig.11 In an exemplary processing flow of , operations include: performing stress analysis 1108 on a portion of the acquired MRI data. In some embodiments, regions of interest can be marked by creating locally induced magnetization perturbations using radio frequency saturation panes. When saturation pulses are applied in two orthogonal planes, the resulting marking pattern forms a grid of intrinsic tissue markers called tags that may deform during contraction. Cardiac stress can be assessed by observing the deformation of the tags.

[0070] As described above, the MRI data acquired during the MRI data acquisition 802 may be reduced before being provided to the algorithm 304 for predicting one or more cardiac signals 806 from the MRI acquisition data, thereby reducing the size of the data to be processed by the algorithm 304. The reduction of the acquired MRI data may also reduce the computation time of the processing engine 306 when performing cine image reconstruction 1208.

[0071] exist Fig.12 In the exemplary processing flow of , operations include: performing cine image reconstruction 1108 on a portion of the acquired MRI data. As described above, the MRI data acquired during MRI data acquisition 802 can be reduced before being provided to the algorithm 304 for predicting one or more cardiac signals 806 from the MRI acquisition data, thereby reducing the size of the data to be processed by the algorithm 304. The reduction of the acquired MRI data can also reduce the computation time of the processing engine 306 when performing cine image reconstruction 1208.

[0072] Exemplary MR image reconstruction techniques may include: using the predicted cardiac cycle and phase information to perform image reconstruction using one or more of a 2D Fourier transform technique, a back projection technique (e.g., a convolution back projection technique, a filtered back projection technique), an iterative reconstruction technique, etc. Exemplary iterative reconstruction techniques may include an algebraic reconstruction technique (ART), a simultaneous iterative reconstruction technique (SIRT), a simultaneous algebraic reconstruction technique (SART), an adaptive statistical iterative reconstruction (ASIR) technique, a model-based iterative reconstruction (MBIR) technique, a sinusoidal pattern confirmation iterative reconstruction (SAFIR) technique, etc., or any combination thereof.

[0073] Go to Fig.13 In an exemplary processing flow, MRI data may be collected from any suitable MRI data source (e.g., one or more of MRI data sources 502, 504, 506, 508, and 510). In addition, ECG signals from the subject may also be collected, which correspond to the collected MRI data. The collected MRI data and the collected ECG data are provided to a deep learning model 308, which operates to predict the cardiac cycle based on the collected MRI data. Operations are performed on a portion of the collected MRI data corresponding to a selected portion of the predicted cardiac cycle portion. Fig.13In the exemplary processing flow of , operations include: performing cine image reconstruction 1308 on a portion of the acquired MRI data. In some embodiments, the MRI data acquired in MRI data acquisition 802 may be reduced before being provided to the algorithm 304 for predicting one or more cardiac signals 1306 from ECG signals and MRI acquisition data, thereby reducing the size of data to be processed by the algorithm 304. The reduction of the acquired MRI data may also reduce the computation time of the processing engine 306 when performing cine image reconstruction 1308 on a selected portion of the MRI data.

[0074] When utilizing the optional ECG signal, Fig.13 The processing flow and Figures 8 to 12 The processing flow is advantageous in cases where cardiac function is impaired and the cardiac ECG signal is adversely affected by imperfect corresponding cardiac motion, structure, blood flow or other cardiac characteristics. For these cases, the deep learning model can use the adversely affected ECG signal and the acquired MRI data to more accurately predict the cardiac cycle. Although the ECG signal may be adversely affected, it can still provide useful information to the deep learning model to more accurately predict the cardiac cycle and cardiac phase. It should be understood that although in Figures 8 to 12 In the workflow of , the use of the ECG signal is optional, but the MRI scanner 402 collects the ECG signal according to the operation, and as a result, the ECG signal is available for use whether or not it is utilized.

[0075] Thus, although the basic novel features of the present invention as applied to the exemplary embodiments of the present invention have been shown, described and pointed out, it should be understood that those skilled in the art may make various omissions, substitutions and changes in the form and details of the exemplified devices and methods and their operation without departing from the spirit and scope of the presently disclosed invention. Further, it is expressly contemplated that all combinations of those elements that perform substantially the same functions in substantially the same manner to achieve the same results are within the scope of the present invention. Moreover, it should be recognized that the structures and / or elements shown and / or described in conjunction with any disclosed form or embodiment of the present invention may be incorporated into any other disclosed or described or suggested form or embodiment as a general matter of design choice. Therefore, the present invention is intended to be limited only by the scope indicated by the scope of the appended claims.

Claims

1. A cardiac signal prediction method, comprising: Acquiring MRI data; wherein the acquired MRI data includes video images from the subject acquired by a camera during an MRI scan; predicting a cardiac cycle from the acquired MRI data using an algorithm; and performing an operation on a portion of the acquired MRI data corresponding to the selected portion of the predicted cardiac cycle; Wherein, irrelevant or unnecessary cardiac cycles are removed from the acquired MRI data before being provided to the algorithm.

2. The method according to claim 1, wherein: The acquired MRI data includes one or more of k-space data, image data, or under-sampled MRI data.

3. The method according to claim 1, wherein: The acquired MRI data includes one or more of an ECG signal or pulse data from a subject acquired during an MRI scan.

4. The method according to claim 1, wherein: The algorithm includes a deep learning model, which further includes a combination of CNN and RNN models, a GRU model, an LSTM model, a fully convolutional neural network model, a generative adversarial network, a back-propagation neural network model, a radial basis function neural network model, a deep belief network neural network model, and one or more of an Elman neural network model.

5. The method according to claim 1, wherein: Operating on the portion of the acquired MRI data corresponding to the selected portion of the predicted cardiac cycle includes: locating data rows in the k-space of the acquired MRI data, or interpolating between MRI data rows in the k-space of the acquired MRI data, or interpolating between MRI images of the acquired MRI data, or using the portion of the acquired MRI data to perform cardiac stress analysis.

6. The method according to claim 1, wherein: Operating on the portion of the acquired MRI data corresponding to the selected portion of the predicted cardiac cycle includes performing cine image reconstruction on the portion of the acquired MRI data.

7. The method according to claim 1, further comprising: Acquiring a cardiac signal corresponding to the MRI data; as well as using the algorithm to predict one or more predicted cardiac signals from the MRI data and the acquired cardiac signals, Wherein, operating the portion of the acquired MRI data corresponding to the selected portion of the predicted cardiac cycle includes: performing cine image reconstruction on the portion of the acquired MRI data.

8. The method according to claim 1, wherein: The predicted portion of the cardiac cycle refers to any portion of the cardiac cycle.

9. The method according to claim 1, wherein: The predicted portion of the cardiac cycle represents one or more of the end-systolic cardiac phase, the end-diastolic cardiac phase, the P, Q, R, S, T, U, QRS complex or the PR interval cardiac phase.

10. A cardiac signal prediction system, comprising: receiving and controlling circuitry configured to operate as an algorithm for predicting cardiac cycles from MRI data, wherein the MRI data includes video images of a subject acquired by a camera during an MRI scan; and a processing engine configured to operate on a portion of the MRI data corresponding to a selected portion of the predicted cardiac cycle; Therein, irrelevant or unnecessary cardiac cycles are removed from the MRI data before being provided to the algorithm.

Images