Method and device for generating heartbeat video based on electromagnetic signal and electronic equipment

Abstract

The invention provides a method and a device for generating a heartbeat video based on an electromagnetic signal, and electronic equipment, which can be applied to the technical field of electromagnetic induction. The method for generating the heartbeat video based on the electromagnetic signal comprises the steps that a target electromagnetic echo signal corresponding to a heart area of a to-be-detected object is determined, and the target electromagnetic echo signal comprises heartbeat motion characteristics after respiratory motion characteristics are eliminated; the target electromagnetic echo signal is subjected to feature mapping processing through an encoder of an electromechanical mode transformation model, target mechanical motion state features of the heart area are obtained, and the target mechanical motion state features correspond to the features of the electrocardio motion state of the to-be-detected object; the target mechanical motion state features and the condition image are processed through an echocardiography video sub-model of the electromechanical mode transformation model, a target echocardiography video of the to-be-detected object is generated, and the condition image represents an echocardiography template image.

Description

Technical Field

[0001] The present invention relates to the field of electromagnetic induction technology, and more specifically to a method, device and electronic equipment for generating a cardiac video based on an electromagnetic signal. Background Art

[0002] Ultrasonic cardiography is one of the most important biomedical images. Since the process of obtaining the target's ultrasonic cardiography using ultrasonic equipment has the advantages of being non-invasive and highly safe, it is widely used in medical testing to facilitate targeted diagnosis of the above-mentioned target as an intermediate result. Furthermore, since the ultrasonic probe is large in size and needs to be connected to a large control device, the acquisition scenario of ultrasonic cardiography is limited, so scientists use electromagnetic signals to monitor the target's cardiac activity.

[0003] In the process of realizing the above-mentioned inventive concept, it was found that: in the related technology, since only some parameters related to the heart can be obtained from the electromagnetic signal, it is difficult to obtain a complete and accurate image of the heart activity. Furthermore, since the floating of the chest surface caused by the heart beating is relatively weak, it is difficult to obtain high-precision heartbeat characteristics from the electromagnetic signal, and it is difficult to achieve the technical problem of accurate conversion of body surface movement-heart movement. Summary of the invention

[0004] In view of the above problems, the present invention provides a method, device and electronic device for generating a cardiac video based on an electromagnetic signal.

[0005] According to a first aspect of the present invention, a method for generating a cardiac video based on an electromagnetic signal is provided, comprising: determining a target electromagnetic echo signal corresponding to a cardiac region of an object to be measured, wherein the target electromagnetic echo signal comprises cardiac motion characteristics after eliminating respiratory motion characteristics; performing feature mapping processing on the target electromagnetic echo signal using an encoder of an electromechanical modal conversion model to obtain target mechanical motion state characteristics of the cardiac region, wherein the target mechanical motion state characteristics correspond to characteristics of the electrocardiographic motion state of the object to be measured; processing the target mechanical motion state characteristics and a conditional image using an ultrasonic cardiac video sub-model of the electromechanical modal conversion model to generate a target ultrasonic cardiac video of the object to be measured, wherein the conditional image represents an ultrasonic cardiac template image.

[0006] According to an embodiment of the present invention, determining a target electromagnetic echo signal corresponding to a heart region of an object to be measured includes: acquiring a plurality of initial electromagnetic echo signals of the object to be measured, wherein the initial electromagnetic echo signal includes a plurality of signals spatially overlapped with each other and reflected by a plurality of body regions of the object to be measured within a predetermined time period; performing fast Fourier transform on the plurality of initial electromagnetic echo signals to determine a two-way propagation distance of each initial electromagnetic echo signal, wherein the two-way propagation distance characterizes a propagation distance of a transmitted signal from a transmitting antenna to a body region and a propagation distance of the initial electromagnetic echo signal reflected from the body region to a receiving antenna; determining a target two-way propagation distance from the plurality of two-way propagation distances based on position information of the object to be measured; and processing each initial electromagnetic echo signal corresponding to the target two-way propagation distance to obtain a target electromagnetic echo signal.

[0007] According to an embodiment of the present invention, each initial electromagnetic echo signal corresponding to a target round-trip propagation distance is processed to obtain a target electromagnetic echo signal, including: determining each initial electromagnetic echo signal corresponding to the target round-trip propagation distance as a plurality of intermediate electromagnetic echo signals; performing fast Fourier transform on the plurality of intermediate electromagnetic echo signals to determine a propagation angle of each intermediate electromagnetic echo signal, wherein the propagation angle is characterized as an angle between the intermediate electromagnetic echo signal and an object to be measured; determining a target propagation angle from a plurality of propagation angles based on angle information of the object to be measured relative to a receiving antenna; and performing second-order differential filtering on the intermediate electromagnetic echo signal corresponding to the target propagation angle to obtain a target electromagnetic echo signal.

[0008] According to an embodiment of the present invention, an ultrasonic video sub-model of an electromechanical modal conversion model is used to process target mechanical motion state characteristics and conditional images to generate a target ultrasonic video of the object to be measured, including: obtaining a Gaussian noise matrix for the ultrasonic video; inputting the Gaussian noise matrix, the target mechanical motion state characteristics and the conditional image into the ultrasonic video sub-model; based on the target mechanical motion state characteristics and the conditional image, the ultrasonic video sub-model denoises the Gaussian noise matrix to generate a target ultrasonic video of the object to be measured.

[0009] According to an embodiment of the present invention, the electromechanical modal conversion model is trained in the following manner, including: obtaining an electromechanical modal conversion model to be trained and a training sample data set, wherein the electromechanical modal conversion model to be trained includes an encoder and an ultrasonic video sub-model to be trained, and the training sample data set includes electromagnetic echo signal samples, ultrasonic electrocardiogram signal samples, ultrasonic video samples and Gaussian noise samples; inputting the ultrasonic electrocardiogram signal samples into the encoder for feature mapping processing, and outputting electrocardiogram motion state feature samples; inputting the electrocardiogram motion state feature samples and ultrasonic video samples into the ultrasonic video sub-model to be trained for processing, and generating training ultrasonic video noise features; based on a video noise loss function, obtaining a video noise loss value according to the training ultrasonic video noise features and the Gaussian noise samples; adjusting the model parameters of the ultrasonic video sub-model to be trained according to the video noise loss value, and obtaining a trained electromechanical modal conversion model.

[0010] According to an embodiment of the present invention, the encoder of the electromechanical modal conversion model to be trained is trained in the following manner, including: obtaining an initial encoder to be trained and an ECG decoder to be trained, wherein the initial encoder includes an electromagnetic encoder to be trained and an ECG encoder to be trained; inputting electromagnetic echo signal samples into the electromagnetic encoder to be trained for feature mapping processing to generate mechanical motion state feature samples; inputting ultrasonic ECG signal samples into the ECG encoder to be trained for feature mapping processing to obtain ECG motion state feature samples; inputting ECG motion state feature samples into the ECG decoder to be trained for signal reconstruction processing to obtain ultrasonic ECG reconstruction signals; obtaining a modal conversion loss value according to a predetermined modal conversion loss function, mechanical motion state feature samples, ECG motion state feature samples, ultrasonic ECG reconstruction signals and ultrasonic ECG signal samples; adjusting model parameters of the initial encoder to be trained and the ECG decoder to be trained according to the modal conversion loss value to obtain a trained encoder and ECG decoder; and determining the trained encoder as the encoder of the electromechanical modal conversion model to be trained.

[0011] According to an embodiment of the present invention, a modal conversion loss value is obtained according to a predetermined modal conversion loss function, mechanical motion state feature samples, ECG motion state feature samples, ultrasonic ECG reconstruction signals and ultrasonic ECG signal samples, including: obtaining motion state feature similarity according to a predetermined motion state feature similarity function, mechanical motion state feature samples and ECG motion state feature samples; obtaining reconstructed signal feature similarity according to a predetermined reconstructed signal feature similarity function, ultrasonic ECG reconstruction signals and ultrasonic ECG signal samples; obtaining a modal conversion loss value according to a predetermined modal conversion loss function, balance parameters, motion state feature similarity and reconstructed signal feature similarity.

[0012] According to an embodiment of the present invention, electrocardiographic motion state feature samples and ultrasonic cardiography video samples are input into an ultrasonic cardiography video sub-model to be trained for processing, and training ultrasonic video noise features are output, including: adding Gaussian noise samples to ultrasonic cardiography video samples to obtain ultrasonic cardiography noise video; inputting ultrasonic cardiography noise video, electrocardiography motion state feature samples and ultrasonic cardiography video samples into an ultrasonic cardiography video sub-model to be trained for denoising processing, and outputting training ultrasonic video noise features.

[0013] The second aspect of the present invention provides a device for generating a cardiac video based on an electromagnetic signal, the device comprising: a determination module, used to determine a target electromagnetic echo signal corresponding to the heart area of ​​a subject to be measured, wherein the target electromagnetic echo signal comprises a heartbeat motion feature after eliminating the respiratory motion feature; a feature mapping module, used to perform feature mapping processing on the target electromagnetic echo signal using an encoder of an electromechanical modal conversion model to obtain a target mechanical motion state feature of the heart area, wherein the target mechanical motion state feature corresponds to the feature of the electrocardiographic motion state of the subject to be measured; a generation module, used to process the target mechanical motion state feature and a conditional image using an ultrasonic cardiac video sub-model of the electromechanical modal conversion model to generate a target ultrasonic cardiac video of the subject to be measured, wherein the conditional image represents an ultrasonic cardiac template image.

[0014] A third aspect of the present invention provides an electronic device, comprising: one or more processors; a memory for storing one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors execute the above method.

[0015] The fourth aspect of the present invention also provides a computer-readable storage medium having executable instructions stored thereon, which, when executed by a processor, causes the processor to execute the above method.

[0016] The fifth aspect of the present invention also provides a computer program product, including a computer program, which implements the above method when executed by a processor.

[0017] According to the method, device and electronic device for generating cardiac video based on electromagnetic signal of the present invention, by determining the target electromagnetic echo signal corresponding to the cardiac region of the object to be measured, the encoder of the electromechanical modal conversion model is used to perform feature mapping processing on the target electromagnetic echo signal, the target mechanical motion state characteristics of the cardiac region are obtained, and the ultrasonic cardiography video sub-model of the electromechanical modal conversion model is used to process the target mechanical motion state characteristics and conditional images, and the target ultrasonic cardiography video of the object to be measured is generated. It is achieved that there is no need to use ultrasonic equipment limited by scenes, and non-contact radar equipment can be used to perform high-granularity measurement on the object to be measured, and the received electromagnetic echo signal is extracted. Based on the electromagnetic echo signal corresponding only to the cardiac region, the high-precision mechanical motion state characteristics of the heartbeat of the object to be measured can be obtained, thereby realizing the accurate conversion of the body surface motion of the object to be measured to the cardiac motion, obtaining the relevant features that can be used to generate the cardiac activity image, and generating the accurate and complete ultrasonic cardiography video of the object to be measured. Without contacting the object to be measured, the monitoring experience of the object to be measured is improved, and it is easier for professionals and the object to be measured to perform convenient operations.

[0018] According to the embodiment of the present invention, further, an electromechanical modal conversion model is constructed based on the encoder and the ultrasonic cardiography video sub-model, and a large amount of model training is performed on the constructed electromechanical modal conversion model. By using the encoder of the pre-trained electromechanical modal conversion model to process the target electromagnetic echo signal, a highly fine-grained target mechanical motion state feature of the heart of the object to be measured is obtained. Since the target mechanical motion state feature extracted by the encoder of the model is similar to the feature of the electrocardiographic motion state of the object to be measured, the target mechanical motion state feature extracted by the encoder can be infinitely close to the real cardiac mechanical activity feature of the object to be measured. In order to input the target mechanical motion state feature as the feature of the electrocardiographic motion state into the ultrasonic cardiography video sub-model of the pre-trained electromechanical modal conversion model, the relevant video generation processing is performed on the pre-prepared ultrasonic cardiography template image of the heart according to the target mechanical motion state feature, thereby improving the processing efficiency in the application process and reducing the working cost. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] The above contents and other objects, features and advantages of the present invention will become more apparent through the following description of the embodiments of the present invention with reference to the accompanying drawings, in which:

[0020] Figure 1 A diagram showing an application scenario of a method for generating a cardiac video based on an electromagnetic signal according to an embodiment of the present invention;

[0021] Figure 2 A flow chart showing a method for generating a cardiac video based on an electromagnetic signal according to an embodiment of the present invention is shown;

[0022] Figure 3a A schematic diagram showing an actual echocardiogram according to an embodiment of the present invention is shown;

[0023] Figure 3b A schematic diagram showing a radar image of a heart obtained by using the method of the present invention according to an embodiment of the present invention is shown;

[0024] Figure 4 A schematic diagram showing a comparison between a partial mechanical motion state characteristic obtained according to an embodiment of the present invention and a left ventricular volume obtained by an actual partial cardiac motion state;

[0025] Figure 5 A schematic diagram showing a comparison between the whole mechanical motion state characteristics obtained according to the present invention and the left ventricular volume obtained according to the actual whole cardiac motion state according to an embodiment of the present invention;

[0026] Figure 6 A structural block diagram of a device for generating a cardiac video based on an electromagnetic signal according to an embodiment of the present invention is shown;

[0027] Figure 7 A block diagram of an electronic device for a method of generating a cardiac video based on an electromagnetic signal according to an embodiment of the present invention is shown. DETAILED DESCRIPTION

[0028] Below, embodiments of the present invention will be described with reference to the accompanying drawings. However, it should be understood that these descriptions are exemplary only and are not intended to limit the scope of the present invention. In the following detailed description, for ease of explanation, many specific details are set forth to provide a comprehensive understanding of embodiments of the present invention. However, it is apparent that one or more embodiments may also be implemented without these specific details. In addition, in the following description, descriptions of known structures and technologies are omitted to avoid unnecessary confusion of concepts of the present invention.

[0029] The terms used herein are only for describing specific embodiments and are not intended to limit the present invention. The terms "comprise", "include", etc. used herein indicate the existence of the features, steps, operations and / or components, but do not exclude the existence or addition of one or more other features, steps, operations or components.

[0030] All terms (including technical and scientific terms) used herein have the meanings commonly understood by those skilled in the art unless otherwise defined. It should be noted that the terms used herein should be interpreted as having a meaning consistent with the context of this specification and should not be interpreted in an idealized or overly rigid manner.

[0031] When using expressions such as "at least one of A, B, and C, etc.", they should generally be interpreted according to the meaning of the expression commonly understood by those skilled in the art (for example, "a system having at least one of A, B, and C" should include but is not limited to a system having A alone, B alone, C alone, A and B, A and C, B and C, and / or A, B, C, etc.).

[0032] In the technical solution of the present invention, the user information (including but not limited to user personal information, user image information, user device information, such as location information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved are all information and data authorized by the user or fully authorized by all parties, and the collection, storage, use, processing, transmission, provision, disclosure and application of the relevant data comply with relevant laws, regulations and standards, take necessary confidentiality measures, do not violate public order and good morals, and provide corresponding operation entrances for users to choose to authorize or refuse.

[0033] As a non-invasive, safe and effective method, ultrasound video is widely used in medical testing to help with targeted diagnosis of the target as an intermediate result. By observing the propagation of ultrasound waves in the human body, ultrasound video can show the movement of the heart, thereby providing video image information that can assist in the diagnosis of the disease. However, the acquisition process of ultrasound video is limited by the standardization of the operator's acquisition process, the variability of the instrument, etc., and the common ultrasound probe is large in size and needs to be connected to a large control device, which greatly limits the acquisition scenario of ultrasound video.

[0034] On this basis, physiological activity monitoring based on electromagnetic signals has the advantages of dynamic monitoring, non-invasiveness and ease of use, and is therefore widely used in monitoring fields such as home health monitoring, remote patient monitoring and public health monitoring, so as to help relevant professionals make auxiliary diagnoses. During the research and development process, it was found that the relevant technology can only obtain some parameters related to the heart from the electromagnetic signal, which makes it difficult to obtain a complete and accurate image of the heart activity. Furthermore, the floating of the chest surface caused by the heartbeat is relatively weak, which makes it difficult to obtain high-precision heartbeat characteristics from the electromagnetic signal, and thus it is difficult to achieve the technical problem of accurate conversion of body surface movement-heart movement.

[0035] In view of this, an embodiment of the present invention provides a method for generating a cardiac video based on an electromagnetic signal, comprising: determining a target electromagnetic echo signal corresponding to the heart region of an object to be measured, wherein the target electromagnetic echo signal includes heartbeat motion characteristics after eliminating respiratory motion characteristics; performing feature mapping processing on the target electromagnetic echo signal using an encoder of an electromechanical modal conversion model to obtain target mechanical motion state characteristics of the heart region, wherein the target mechanical motion state characteristics correspond to characteristics of the electrocardiographic motion state of the object to be measured; processing the target mechanical motion state characteristics and a conditional image using an ultrasonic cardiac video sub-model of the electromechanical modal conversion model to generate a target ultrasonic cardiac video of the object to be measured, wherein the conditional image represents an ultrasonic cardiac template image.

[0036] Figure 1 A diagram showing an application scenario of a method for generating a cardiac video based on an electromagnetic signal according to an embodiment of the present invention.

[0037] like Figure 1 As shown, the application scenario according to this embodiment may include a first radar device 101, a target object 102, and a receiver 103. The first radar device 101 is used to send an electromagnetic wave signal to the target object 102.

[0038] The user may use the first radar device 101 to interact with the target object 102 and the receiver 103 to receive or send signals, etc.

[0039] The receiver 103 may be a receiver for receiving various electromagnetic echo signals, for example, receiving and processing the electromagnetic signal sent by the first radar device 101 (for example only). The receiver 103 may analyze and process the received electromagnetic echo signal and other data, and feed back the processing result (for example, obtaining or generating relevant measurement data according to a user request, etc.) to the terminal device.

[0040] It should be noted that the method for generating a cardiac video based on an electromagnetic signal provided in the embodiment of the present invention can generally be performed by the receiver 103. Accordingly, the device for generating a cardiac video based on an electromagnetic signal provided in the embodiment of the present invention can generally be set in the receiver 103. The method for generating a cardiac video based on an electromagnetic signal provided in the embodiment of the present invention can also be performed by a receiver or a receiver cluster that is different from the receiver 103 and can communicate with the first radar device 101 and / or the receiver 103. Accordingly, the device for generating a cardiac video based on an electromagnetic signal provided in the embodiment of the present invention can also be set in a receiver or a receiver cluster that is different from the receiver 103 and can communicate with the first radar device 101 and / or the receiver 103.

[0041] It should be understood that Figure 1The number of the first radar device, the target object and the receiver in the embodiment is only for illustration. According to the implementation requirements, there may be any number of radar devices, observation objects and receivers.

[0042] The following will be based on Figure 1 The scene described by Figure 2~Figure 5 The method for generating a cardiac video based on an electromagnetic signal according to an embodiment of the present invention is described in detail.

[0043] Figure 2 A flow chart of a method for generating a cardiac video based on an electromagnetic signal according to an embodiment of the present invention is shown.

[0044] like Figure 2 As shown, the method for generating a cardiac video based on an electromagnetic signal in this embodiment includes operations S210 to S230.

[0045] In operation S210, a target electromagnetic echo signal corresponding to a heart region of the subject to be measured is determined.

[0046] According to an embodiment of the present invention, the target electromagnetic echo signal includes a heartbeat motion feature after eliminating the respiratory motion feature.

[0047] According to an embodiment of the present invention, an electromagnetic signal is transmitted to multiple body regions of the object to be measured by a transmitting antenna. When receiving the electromagnetic signal, the multiple body regions of the object to be measured generate electromagnetic echo signals and reflect the electromagnetic echo signals to the receiving antenna, so that a target electromagnetic echo signal corresponding to the heart region of the object to be measured can be determined from the multiple electromagnetic echo signals, so that the heart of the object to be measured can be processed according to the target electromagnetic echo signal corresponding to the heart region to obtain an accurate ultrasonic cardiography video.

[0048] In operation S220, the target electromagnetic echo signal is subjected to feature mapping processing using an encoder of the electromechanical mode conversion model to obtain target mechanical motion state features of the heart region.

[0049] According to an embodiment of the present invention, the target mechanical motion state characteristics correspond to the characteristics of the electrocardiographic motion state of the object to be measured, that is, the target mechanical motion state characteristics are similar to the characteristics of the electrocardiographic motion state, so that both characteristics are infinitely close to the real cardiac mechanical activity characteristics.

[0050] According to an embodiment of the present invention, an encoder of a pre-trained electromechanical modal conversion model is used to extract and map features in an acquired accurate target electromagnetic echo signal to obtain target mechanical motion state features of the heart region of the subject to be measured.

[0051] In operation S230, the target mechanical motion state characteristics and the conditional image are processed using the ultrasound video sub-model of the electromechanical mode conversion model to generate a target ultrasound video of the object to be measured.

[0052] According to an embodiment of the present invention, the conditional image may represent an echocardiography template image.

[0053] According to an embodiment of the present invention, when the template image database contains historical ultrasonic images of the object to be measured, the historical ultrasonic images of the object to be measured can be input as conditional images into the ultrasonic video sub-model of the electromechanical modal conversion model to generate a target ultrasonic video of the object to be measured; when the template image database does not contain historical ultrasonic images of the object to be measured, the ultrasonic template image can be input into the ultrasonic video sub-model of the electromechanical modal conversion model to generate a target ultrasonic video of the object to be measured.

[0054] According to the embodiment of the present invention, the obtained target ultrasound cardiology video can help relevant personnel to judge the relevant functions of the target.

[0055] For example, an electromagnetic signal is transmitted to the heart of the target object, a plurality of electromagnetic echo signals are received by a receiving antenna, a target electromagnetic echo signal corresponding to the heart region of the target object is determined from the plurality of electromagnetic echo signals, a feature mapping process is performed on the target electromagnetic echo signal by using an encoder of a trained electromechanical modal conversion model, the mechanical motion features of the heart in the target electromagnetic echo signal of the target object are extracted to obtain target mechanical motion state features, and then the target mechanical motion state features and the ultrasonic cardiography template image are input into the ultrasonic cardiography video sub-model of the trained electromechanical modal conversion model for processing to generate a target ultrasonic cardiography video of the target object so that scientists or professionals in related fields can identify the cardiac function of the target object based on the target ultrasonic cardiography video of the target object, thereby making a judgment appropriate for the target object.

[0056] According to an embodiment of the present invention, by determining the target electromagnetic echo signal corresponding to the heart region of the object to be measured, the target electromagnetic echo signal is feature mapped using the encoder of the electromechanical modal conversion model to obtain the target mechanical motion state characteristics of the heart region, and the target mechanical motion state characteristics and conditional images are processed using the ultrasonic cardiography video sub-model of the electromechanical modal conversion model to generate the target ultrasonic cardiography video of the object to be measured. It is achieved that there is no need to use ultrasonic equipment limited by the scene, and the object to be measured can be measured with high granularity using non-contact radar equipment, and the received electromagnetic echo signal is extracted. Based on the electromagnetic echo signal corresponding only to the heart region, the high-precision heartbeat mechanical motion state characteristics of the object to be measured can be obtained, thereby achieving accurate conversion of the body surface motion of the object to be measured to the heart motion, obtaining relevant features that can be used to generate cardiac activity images, and generating accurate and complete ultrasonic cardiography videos of the object to be measured. Without contacting the object to be measured, the monitoring experience of the object to be measured is improved, and it is easier for professionals and the object to be measured to perform convenient operations.

[0057] According to the embodiment of the present invention, further, an electromechanical modal conversion model is constructed based on the encoder and the ultrasonic cardiography video sub-model, and a large amount of model training is performed on the constructed electromechanical modal conversion model. By using the encoder of the pre-trained electromechanical modal conversion model to process the target electromagnetic echo signal, a highly fine-grained target mechanical motion state feature of the heart of the object to be measured is obtained. Since the target mechanical motion state feature extracted by the encoder of the model is similar to the feature of the electrocardiographic motion state of the object to be measured, the target mechanical motion state feature extracted by the encoder can be infinitely close to the real cardiac mechanical activity feature of the object to be measured. In order to input the target mechanical motion state feature as the feature of the electrocardiographic motion state into the ultrasonic cardiography video sub-model of the pre-trained electromechanical modal conversion model, the relevant video generation processing is performed on the pre-prepared ultrasonic cardiography template image of the heart according to the target mechanical motion state feature, thereby improving the processing efficiency in the application process and reducing the working cost.

[0058] It should be noted that the target ultrasound video obtained by the present invention is only an intermediate result, and the diagnosis result or health condition cannot be directly derived from the target ultrasound video obtained by the method of the present invention.

[0059] According to an embodiment of the present invention, a method for determining a target electromagnetic echo signal corresponding to a heart region of a subject to be measured includes the following operations.

[0060] According to an embodiment of the present invention, a plurality of initial electromagnetic echo signals of the object to be measured are acquired.

[0061] According to an embodiment of the present invention, the initial electromagnetic echo signal includes a plurality of signals that are reflected by a plurality of body regions of the subject to be measured within a predetermined period of time and are spatially overlapped with each other.

[0062] According to an embodiment of the present invention, an electromagnetic signal is transmitted to multiple body regions of the object to be measured using a transmitting antenna, so that the multiple body regions generate multiple electromagnetic echo signals that are spatially overlapped with each other after receiving the electromagnetic signal, and reflect the multiple electromagnetic echo signals that are spatially overlapped with each other to the receiving antenna.

[0063] According to an embodiment of the present invention, a plurality of initial electromagnetic echo signals can be obtained according to formula (1).

[0064] (1);

[0065] Among them, S(x, y, z, m) can be represented as the initial electromagnetic echo signal, N can be represented as the total number of channels composed of transmitting antennas and receiving antennas for transmitting signals, n can be represented as the nth channel, M can be represented as the total duration of receiving electromagnetic echo signals, m can be represented as the tth moment, y n,t It can be represented as the electromagnetic echo signal received by the nth channel at the tth time, and k can be represented as the frequency modulation slope. can be represented as the wavelength of the RF signal, (x,y,z) can be represented as the coordinates in three-dimensional space, and r(x,y,z,n) can be represented as the two-way propagation distance of the electromagnetic echo signal reflected to (x,y,z) in the nth channel.

[0066] According to an embodiment of the present invention, a plurality of initial electromagnetic echo signals are subjected to fast Fourier transform to determine a two-way propagation distance of each initial electromagnetic echo signal.

[0067] According to an embodiment of the present invention, the two-way propagation distance represents the propagation distance of the transmission signal from the transmission antenna to the body region and the propagation distance of the initial electromagnetic echo signal reflected from the body region to the receiving antenna.

[0068] According to an embodiment of the present invention, there are multiple channels for transmitting electromagnetic signals and initial electromagnetic echo signals, and the round-trip propagation distance may be the sum of the round-trip transmission distances of the radar signal in any channel.

[0069] According to an embodiment of the present invention, a plurality of initial electromagnetic echo signals may have the same two-way propagation distance.

[0070] According to an embodiment of the present invention, a target two-way propagation distance is determined from a plurality of two-way propagation distances based on the position information of the object to be measured.

[0071] According to an embodiment of the present invention, the position information of the chest area of ​​the object to be measured can be obtained, and the target two-way propagation distance can be determined based on the position information of the object to be measured. For example, if the distance between the chest area of ​​the object to be measured and the radar antenna is 1m, the target two-way propagation distance is determined to be 2m.

[0072] According to an embodiment of the present invention, each initial electromagnetic echo signal corresponding to the target round-trip propagation distance is processed to obtain a target electromagnetic echo signal.

[0073] According to an embodiment of the present invention, multiple initial electromagnetic echo signals corresponding to the target two-way distance are determined according to the target two-way distance, and the determined initial electromagnetic echo signals are processed to obtain the target electromagnetic echo signal. For example, if the target two-way propagation distance is 2m, the two-way propagation distance of the first initial electromagnetic echo signal in the channel is 3m, the two-way propagation distance of the second initial electromagnetic echo signal in the channel is 2m, the two-way propagation distance of the third initial electromagnetic echo signal in the channel is 9m, and the two-way propagation distance of the fourth initial electromagnetic echo signal in the channel is 2m, then from all the initial electromagnetic echo signals, the second initial electromagnetic echo signal and the fourth initial electromagnetic echo signal corresponding to the target two-way propagation distance are determined to be processed to obtain the target electromagnetic echo signal.

[0074] According to an embodiment of the present invention, multiple initial electromagnetic echo signals of an object to be measured are acquired, fast Fourier transform is performed on the multiple initial electromagnetic echo signals, the round-trip propagation distance of each initial electromagnetic echo signal is determined, the target round-trip propagation distance is determined from the multiple round-trip propagation distances based on the position information of the object to be measured, each initial electromagnetic echo signal corresponding to the target round-trip propagation distance is processed to obtain the target electromagnetic echo signal, thereby obtaining the target electromagnetic echo signal corresponding to the heart area from the initial electromagnetic echo signals corresponding to multiple body areas based on the electromagnetic signal emitted by the radar device and the initial electromagnetic echo signal received by reflection, so as to process the target electromagnetic echo signal corresponding to the heart area, thereby accurately extracting the high-precision heartbeat characteristics of the object to be measured, and then realizing the accurate conversion of body surface movement-heart movement, so as to obtain more accurate and comprehensive partial parameters related to the heart, thereby generating a complete and accurate heart activity image.

[0075] According to an embodiment of the present invention, a method for processing each initial electromagnetic echo signal corresponding to a target two-way propagation distance to obtain a target electromagnetic echo signal includes the following operations.

[0076] According to an embodiment of the present invention, each initial electromagnetic echo signal corresponding to a target round-trip propagation distance is determined as a plurality of intermediate electromagnetic echo signals.

[0077] For example, the target two-way propagation distance is 2m, the two-way propagation distance of the first initial electromagnetic echo signal in the channel is 3m, the two-way propagation distance of the second initial electromagnetic echo signal in the channel is 2m, the two-way propagation distance of the third initial electromagnetic echo signal in the channel is 9m, and the two-way propagation distance of the fourth initial electromagnetic echo signal in the channel is 2m, then from all the initial electromagnetic echo signals, the second initial electromagnetic echo signal and the fourth initial electromagnetic echo signal corresponding to the target two-way propagation distance are determined as intermediate electromagnetic echo signals.

[0078] According to an embodiment of the present invention, a fast Fourier transform is performed on a plurality of intermediate electromagnetic echo signals to determine a propagation angle of each intermediate electromagnetic echo signal.

[0079] According to an embodiment of the present invention, the propagation angle may be characterized as the angle between the intermediate electromagnetic echo signal and the object to be measured.

[0080] According to an embodiment of the present invention, based on the angle information of the object to be measured relative to the receiving antenna, the target propagation angle is determined from multiple propagation angles.

[0081] According to an embodiment of the present invention, an angle close to the angle information of the object to be measured relative to the receiving antenna is determined from multiple propagation angles as the target propagation angle. For example, if the angle of the object to be measured relative to the receiving antenna is 50°, the angle of the first intermediate electromagnetic echo signal is 88°, the angle of the second intermediate electromagnetic echo signal is 51°, and the angle of the third intermediate electromagnetic echo signal is 26°, then the target propagation angle is confirmed to be 51°.

[0082] According to an embodiment of the present invention, a second-order differential filtering process is performed on an intermediate electromagnetic echo signal corresponding to a target propagation angle to obtain a target electromagnetic echo signal.

[0083] For example, the angle of the first intermediate electromagnetic echo signal is 88°, the angle of the second intermediate electromagnetic echo signal is 51°, the angle of the third intermediate electromagnetic echo signal is 26°, and the target propagation angle is 51°. The second intermediate electromagnetic echo signal is subjected to second-order differential filtering to obtain the target electromagnetic echo signal.

[0084] According to the embodiment of the present invention, since the acceleration of respiratory motion is smaller than the acceleration of heart beat, the respiratory motion feature in the intermediate electromagnetic echo signal is removed by second-order differential filtering so as to retain the heart beat motion feature.

[0085] According to an embodiment of the present invention, the target electromagnetic echo signal after second-order differential filtering is as shown in formula (2).

[0086] (2);

[0087] in, can be represented as the target electromagnetic echo signal, h can be represented as the frame cycle time, s0 can be represented as the phase result of the electromagnetic echo signal received at the current moment, s1 can be represented as the phase result of the electromagnetic echo signal received at a moment after the current moment, and s -1 It can be represented as the phase result of the electromagnetic echo signal received at a time before the current time, and s2 can be represented as the phase result of the electromagnetic echo signal received at two times after the current time. -2 It can be represented as the phase result of the electromagnetic echo signal received two times before the current time, and s3 can be represented as the phase result of the electromagnetic echo signal received three times after the current time. -3 It can be characterized as the phase result of the electromagnetic echo signal received three times before the current time.

[0088] According to an embodiment of the present invention, each initial electromagnetic echo signal corresponding to the target two-way propagation distance is determined as a plurality of intermediate electromagnetic echo signals, and a fast Fourier transform is performed on the plurality of intermediate electromagnetic echo signals to determine the propagation angle of each intermediate electromagnetic echo signal. Based on the angle information of the object to be measured relative to the receiving antenna, the target propagation angle is determined from the plurality of propagation angles, and the intermediate electromagnetic echo signals corresponding to the target propagation angle are subjected to second-order differential filtering to obtain the target electromagnetic echo signal, thereby achieving determination of the intermediate electromagnetic echo signal of the heart region of the object to be measured from the plurality of intermediate electromagnetic echo signals, and the determined intermediate electromagnetic echo signal includes respiratory motion features, so that based on the characteristics between respiratory motion and heartbeat motion, the respiratory motion features are filtered out by second-order differential filtering, and the highly fine-grained target electromagnetic echo signal containing the heartbeat motion features is retained, so as to process the highly fine-grained target electromagnetic echo signal to obtain high-precision heartbeat features, thereby achieving accurate conversion of body surface motion-heart motion, so as to obtain more accurate and comprehensive partial parameters related to the heart, thereby generating a complete and accurate heart activity image.

[0089] According to an embodiment of the present invention, multiple initial electromagnetic echo signals can also be processed by beamforming to separate signals that are spatially overlapped with each other, thereby extracting intermediate electromagnetic echo signals of the heart region of the object to be measured, and then performing second-order differential filtering on the intermediate electromagnetic echo signals after beamforming processing to obtain target electromagnetic echo signals.

[0090] According to an embodiment of the present invention, a method for generating a target ultrasonic video of a subject to be tested by processing target mechanical motion state characteristics and conditional images using an ultrasonic video sub-model of an electromechanical modal conversion model includes the following operations.

[0091] According to an embodiment of the present invention, a Gaussian noise matrix for an echocardiography video is obtained.

[0092] According to an embodiment of the present invention, a set of Gaussian noise matrices may be randomly generated, or a set of random Gaussian noise matrices may be obtained from an ultrasonic cardiac noise library.

[0093] According to an embodiment of the present invention, the Gaussian noise matrix can be used together with the target mechanical motion state characteristics and the conditional image to generate a target ultrasound cardiac video of the object to be measured.

[0094] According to an embodiment of the present invention, a Gaussian noise matrix, target mechanical motion state characteristics and conditional images are input into an ultrasonic cardiology video sub-model. Based on the target mechanical motion state characteristics and conditional images, the ultrasonic cardiology video sub-model denoises the Gaussian noise matrix to generate a target ultrasonic cardiology video of the object to be measured.

[0095] According to an embodiment of the present invention, a pre-trained ultrasound cardiology video sub-model is used to perform denoising on a randomly generated Gaussian noise matrix, thereby restoring ultrasound video information without noise. The ultrasound video information without noise is then further processed in combination with the target mechanical motion state characteristics and the ultrasound template image to generate a target ultrasound cardiology video of the object to be measured.

[0096] According to an embodiment of the present invention, a Gaussian noise matrix for an ultrasonic cardiogram video is obtained, and the Gaussian noise matrix, target mechanical motion state characteristics, and conditional images are input into an ultrasonic cardiogram sub-model. Based on the target mechanical motion state characteristics and conditional images, the ultrasonic cardiogram sub-model performs denoising processing on the Gaussian noise matrix to generate a target ultrasonic cardiogram video of the object to be measured. Based on an electromagnetic echo signal, the mechanical motion state of the heart in the electromagnetic echo signal is converted into the electrocardiogram motion state in the ultrasonic video through a pre-trained electromechanical modal conversion model, and an accurate target ultrasonic cardiogram video of the heart region of the object to be measured is generated. An accurate and complete target ultrasonic cardiogram video is obtained while improving the adaptability to application scenarios and facilitating operation. Moreover, since only the electromagnetic echo signal needs to be processed, a variety of signals or video information of the heart region of the object to be measured that can assist in diagnosis can be obtained simultaneously, thereby improving processing efficiency and the experience comfort of the object to be measured.

[0097] According to an embodiment of the present invention, the electromechanical modal conversion model is trained in the following manner.

[0098] According to an embodiment of the present invention, an electromechanical modal conversion model to be trained and a training sample data set are obtained.

[0099] According to an embodiment of the present invention, the electromechanical modal conversion model to be trained includes an encoder and an ultrasound video sub-model to be trained, and the training sample data set includes electromagnetic echo signal samples, ultrasound electrocardiogram signal samples, ultrasound video samples and Gaussian noise samples.

[0100] According to an embodiment of the present invention, the encoder in the electromechanical modal conversion model to be trained may be a pre-trained encoder, and training is performed using the pre-trained encoder and the ultrasonic video sub-model to be trained to obtain a trained ultrasonic video sub-model.

[0101] According to an embodiment of the present invention, the ultrasonic electrocardiogram signal samples are input into an encoder for feature mapping processing to generate electrocardiogram motion state feature samples.

[0102] According to an embodiment of the present invention, characteristic samples of electrocardiographic motion states are extracted and mapped from ultrasonic electrocardiographic signal samples so as to train an ultrasonic cardiography video sub-model to generate accurate ultrasonic cardiography videos.

[0103] According to an embodiment of the present invention, the electrocardiographic motion state feature samples and the ultrasonic cardiography video samples are input into the ultrasonic cardiography video sub-model to be trained for processing, and the training ultrasonic video noise features are output.

[0104] According to an embodiment of the present invention, a method for inputting electrocardiographic motion state feature samples and ultrasonic cardiography video samples into an ultrasonic cardiography video sub-model to be trained for processing and outputting training ultrasonic video noise features includes the following operations.

[0105] According to an embodiment of the present invention, Gaussian noise samples are added to ultrasonic cardiology video samples to obtain ultrasonic cardiology noise video.

[0106] According to an embodiment of the present invention, Gaussian ultrasound samples are randomly acquired, and Gaussian noise samples are added to the ultrasound cardiac video multiple times to obtain multiple ultrasound cardiac noise videos.

[0107] According to an embodiment of the present invention, after the last Gaussian noise sample is added to the ultrasonic cardiac video, the obtained ultrasonic cardiac noise video may be as shown in formula (3).

[0108] (3);

[0109] in, It can be represented as follows: T can be represented as the total number of times Gaussian noise samples are added to the ultrasound video, and t can be represented as the tth time that Gaussian noise samples are added to the ultrasound video. It can be represented as the ultrasonic noise video obtained by adding the t-th Gaussian noise sample to the ultrasonic video after adding the t-1th Gaussian noise sample, x0 can be represented as the ultrasonic video without adding Gaussian noise samples, x 1:T It can be represented as the ultrasonic cardiac noise video obtained by cumulatively adding T Gaussian noise samples to the ultrasonic cardiac video, x t It can be represented as the ultrasonic cardiac noise video obtained by adding the t-th Gaussian noise sample to the ultrasonic cardiac video, x t-1 It can be characterized as an ultrasonic cardiac noise video obtained by adding the t-1th Gaussian noise sample to the ultrasonic cardiac video.

[0110] According to an embodiment of the present invention, after adding the t-1th Gaussian noise sample to the ultrasonic video sample, the ultrasonic noise video obtained by further adding the tth Gaussian noise sample to the ultrasonic video sample can be as shown in formula (4).

[0111] (4);

[0112] in, can be represented as a noise parameter, I can be represented as a unit matrix, It can be characterized as a Gaussian distribution.

[0113] According to an embodiment of the present invention, the ultrasonic cardiology noise video, the electrocardiographic motion state feature samples and the ultrasonic cardiology video samples are input into the ultrasonic cardiology video sub-model to be trained for denoising, and the training ultrasonic video noise features are output.

[0114] According to an embodiment of the present invention, a plurality of ultrasonic cardiac noise videos to which multiple Gaussian noise samples have been added, original ultrasonic cardiac video samples, and electrocardiographic motion state feature samples extracted and mapped by a pre-trained encoder are input into an ultrasonic cardiac video sub-model to be trained. The ultrasonic cardiac video sub-model to be trained repeatedly performs denoising processing on the plurality of ultrasonic cardiac noise videos based on the ultrasonic cardiac video samples and the electrocardiographic motion state features, so as to obtain a plurality of training ultrasonic video noise features.

[0115] According to an embodiment of the present invention, multiple ultrasonic noise videos are obtained by adding Gaussian noise samples to ultrasonic video multiple times, and multiple ultrasonic noise videos, ultrasonic video samples and electrocardiographic motion state feature samples are input into the ultrasonic video sub-model to be trained. The ultrasonic video sub-model to be trained repeatedly performs denoising processing on multiple ultrasonic noise videos based on the ultrasonic video samples and electrocardiographic motion state features, so as to obtain multiple training ultrasonic video noise features, thereby realizing the training of the ultrasonic video sub-model, so as to use the trained ultrasonic video sub-model to process the input data and generate accurate target ultrasonic video. At the same time, by repeatedly performing denoising processing on multiple ultrasonic noise videos, the ultrasonic video sub-model to be trained can learn to obtain the training ultrasonic video noise features that can be applied to the elimination of various Gaussian noises, thereby improving the recognition ability of ultrasonic video noise features, so that in the application process, the Gaussian noise matrix used to generate the ultrasonic video can be denoised based on the target mechanical motion state features and the conditional image, thereby further improving the processing efficiency and generation accuracy.

[0116] According to an embodiment of the present invention, a video noise loss value is obtained according to a video noise loss function, a training ultrasound video noise feature and a Gaussian noise sample.

[0117] According to an embodiment of the present invention, the video noise loss function may be as shown in formula (5).

[0118] (5);

[0119] in, It can be characterized as a video noise loss function, It can be characterized as the Gaussian noise sample added for the tth time, It can be characterized as the noise feature of the training ultrasound video. can be characterized as a noise parameter, z e It can be characterized as a characteristic sample of the electrocardiographic motion state.

[0120] According to an embodiment of the present invention, the model parameters of the ultrasound cardiology video sub-model to be trained are adjusted according to the video noise loss value to obtain a trained electromechanical mode conversion model.

[0121] According to an embodiment of the present invention, when the training of the ultrasound video sub-model of the electromechanical modal conversion model is completed, the ultrasound video sub-model can be based on the standard Gaussian distribution x under given conditional images. T ∼N(0,I) ECG state feature samples are used to restore ultrasound cardiogram video samples.

[0122] According to an embodiment of the present invention, a trained ultrasonic video sub-model is obtained by adjusting the model parameters of the ultrasonic video sub-model to be trained, and a trained electromechanical modal conversion model can be obtained by combining with a pre-trained encoder.

[0123] According to an embodiment of the present invention, a pre-trained encoder is used to perform feature mapping processing on ultrasonic electrocardiogram signal samples in a training sample data set to generate electrocardiogram motion state feature samples, the electrocardiogram motion state feature samples and ultrasonic cardiogram video samples obtained by the encoder are input into an ultrasonic cardiogram video sub-model to be trained, and the training ultrasonic cardiogram noise features are output. According to a video noise loss function, the training ultrasonic cardiogram noise features and Gaussian noise samples, a video noise loss value is obtained, thereby adjusting the model parameters of the ultrasonic cardiogram video sub-model to be trained, and obtaining a trained electromechanical modal conversion model, thereby realizing the electromechanical modal conversion of the ultrasonic cardiogram video sub-model in the electromechanical modal conversion model to be trained according to the electrocardiogram motion state feature samples obtained by processing the pre-trained encoder. The training of the type is carried out to improve the correlation of feature processing between the ultrasonic video sub-model and the pre-trained encoder, so that the trained ultrasonic video sub-model can be used to perform video generation processing according to the features obtained by the encoder, and an accurate ultrasonic video corresponding to the features output by the encoder is obtained. At the same time, the ultrasonic video is also an accurate target ultrasonic video of the object to be measured. Furthermore, by using a large number of data samples in the training sample data set to train the electromechanical modal conversion model to be trained, the robustness of the electromechanical modal conversion model is improved, so that the trained electromechanical modal conversion model can be used to process the features of the actually measured signal to generate an accurate target ultrasonic video, thereby improving the model processing efficiency.

[0124] According to an embodiment of the present invention, the encoder of the electromechanical modal conversion model to be trained is pre-trained.

[0125] According to an embodiment of the present invention, the encoder of the electromechanical modal conversion model to be trained is trained in the following manner.

[0126] According to an embodiment of the present invention, an initial encoder to be trained and an electrocardiogram decoder to be trained are obtained.

[0127] According to an embodiment of the present invention, the initial encoder includes an electromagnetic encoder to be trained and an electrocardiographic encoder to be trained.

[0128] According to an embodiment of the present invention, the electromagnetic encoder to be trained may be a space-time encoder to be trained, wherein the space-time encoder may include a space encoder and a time sequence encoder, and the electrocardiogram encoder to be trained may be a time sequence encoder to be trained.

[0129] According to an embodiment of the present invention, the electromagnetic echo signal samples are input to the electromagnetic encoder to be trained for feature mapping processing to generate mechanical motion state feature samples.

[0130] According to an embodiment of the present invention, the mechanical motion state feature sample may include the temporal features and spatial features of the cardiac mechanical motion.

[0131] According to an embodiment of the present invention, ultrasonic electrocardiogram signal samples are input into an electrocardiogram encoder to be trained for feature mapping processing to obtain electrocardiogram motion state feature samples.

[0132] According to an embodiment of the present invention, the electrocardiographic motion state feature sample may include the timing characteristics of the heart.

[0133] According to an embodiment of the present invention, the similarity between the two features can be first obtained based on the mechanical motion state feature samples and the electrocardiographic motion state feature samples, and the parameters of the electromagnetic encoder to be trained and the electrocardiographic encoder to be trained are adjusted based on the similarity, so as to obtain the electromagnetic encoder and electrocardiographic encoder that have been preliminarily trained.

[0134] According to an embodiment of the present invention, by aligning the mechanical motion state feature samples and the electrocardiographic motion state feature samples, the two features can be infinitely close to the cardiac mechanical activity features, thereby obtaining a high-precision encoder, and further realizing the accurate conversion of body surface motion-heart motion, so that the training results obtained by the encoder can be directly used to train the ultrasound cardiography video sub-model to be trained.

[0135] According to an embodiment of the present invention, the ECG motion state characteristic samples are input into an ECG decoder to be trained for signal reconstruction processing to obtain an ultrasonic ECG reconstruction signal.

[0136] According to an embodiment of the present invention, the ECG motion state characteristic samples processed by the ECG encoder to be trained are input into the ECG decoder to be trained, so that the ECG decoder to be trained can reconstruct the ECG motion state characteristic samples and restore the ultrasonic ECG reconstructed signals similar to the ultrasonic ECG signal samples not processed by the ECG encoder.

[0137] According to an embodiment of the present invention, signal similarity can be obtained based on the ultrasonic ECG reconstruction signal and the ultrasonic ECG signal sample output by the ECG decoder to be trained, and the parameters of the ECG decoder to be trained are adjusted according to the signal similarity to obtain a preliminarily trained ECG decoder.

[0138] According to an embodiment of the present invention, a modal conversion loss value is obtained based on a predetermined modal conversion loss function, mechanical motion state feature samples, electrocardiographic motion state feature samples, ultrasonic electrocardiographic reconstruction signals and ultrasonic electrocardiographic signal samples.

[0139] According to an embodiment of the present invention, the model parameters of the electromagnetic encoder to be trained, the ECG encoder to be trained and the ECG decoder to be trained can be adjusted again according to the obtained modal conversion loss value, so as to obtain an encoder and a decoder that can output accurate features or signals.

[0140] According to an embodiment of the present invention, a method for obtaining a modal conversion loss value according to a predetermined modal conversion loss function, mechanical motion state feature samples, electrocardiographic motion state feature samples, ultrasonic electrocardiographic reconstruction signals and ultrasonic electrocardiographic signal samples includes the following operations.

[0141] According to an embodiment of the present invention, the motion state feature similarity is obtained based on a predetermined motion state feature similarity function, a mechanical motion state feature sample, and an electrocardiogram motion state feature sample.

[0142] According to an embodiment of the present invention, the motion state feature similarity function may be as shown in formula (6), and the motion state feature similarity may be calculated according to formula (6).

[0143] (6);

[0144] Among them, l z It can be represented as the similarity of motion state features, z r It can be characterized as a characteristic sample of mechanical motion state, z e It can be characterized as a characteristic sample of the electrocardiographic motion state.

[0145] According to an embodiment of the present invention, preliminary parameter adjustments can be made to the electromagnetic encoder to be trained and the electrocardiogram encoder to be trained based on the similarity of motion state features, so that the mechanical motion state feature samples generated by the electromagnetic encoder to be trained and the electrocardiogram motion state feature samples generated by the electrocardiogram encoder to be trained have a high feature similarity, so that when the electrocardiogram motion state feature samples are used to train the ultrasonic cardiogram video sub-model to be trained, it is equivalent to using the mechanical motion state feature samples to train the ultrasonic cardiogram video sub-model to be trained, thereby improving the output accuracy of the model and the encoder.

[0146] According to an embodiment of the present invention, the reconstructed signal feature similarity is obtained based on a predetermined reconstructed signal feature similarity function, an ultrasonic electrocardiogram reconstructed signal and an ultrasonic electrocardiogram signal sample.

[0147] According to an embodiment of the present invention, the reconstructed signal feature similarity function may be as shown in formula (7), and the reconstructed signal feature similarity may be calculated according to formula (7).

[0148] (7);

[0149] Among them, l rec It can be represented as the similarity of reconstructed signal features. It can be characterized as an ultrasound ECG reconstruction signal, e i It can be characterized as an ultrasonic electrocardiogram signal sample.

[0150] According to an embodiment of the present invention, preliminary parameter adjustment can be performed on the ECG decoder to be trained based on the similarity of reconstructed signal features, so that the ultrasonic ECG reconstructed signal restored by the ECG decoder to be trained can be consistent with the original ultrasonic ECG signal sample.

[0151] According to an embodiment of the present invention, a modal conversion loss value is obtained according to a predetermined modal conversion loss function, a balance parameter, a motion state feature similarity, and a reconstructed signal feature similarity.

[0152] According to an embodiment of the present invention, the modal conversion loss function may be as shown in formula (8), and the modal conversion loss value may be calculated according to formula (8).

[0153] (8);

[0154] in, It can be represented as the modal conversion loss value, It can be characterized as a balance weight.

[0155] According to an embodiment of the present invention, by obtaining the motion state feature similarity according to a predetermined motion state feature similarity function, a mechanical motion state feature sample and an ECG motion state feature sample, obtaining the reconstructed signal feature similarity according to a predetermined reconstructed signal feature similarity function, an ultrasonic ECG reconstructed signal and an ultrasonic ECG signal sample, and obtaining the modal conversion loss value according to a predetermined modal conversion loss function, a balance parameter, the motion state feature similarity and the reconstructed signal feature similarity, it is possible to implement a parameter adjustment of the electromagnetic encoder to be trained and the ECG encoder to be trained according to the motion state feature similarity, so as to obtain an electromagnetic encoder and an ECG encoder with close feature similarity, and then, according to the reconstructed signal feature similarity, the ECG decoder to be trained is adjusted once, so as to obtain an ECG decoder that can reconstruct an ECG decoder that is consistent with the original ultrasonic ECG sample signal, and finally, based on the modal conversion loss value, the parameter weights of the adjusted electromagnetic encoder, ECG encoder and ECG decoder are adjusted again, so as to obtain an electromagnetic encoder and an ECG encoder that can output high-precision features, thereby achieving accurate conversion of body surface motion-heart motion, and improving the output accuracy and processing efficiency of the encoder.

[0156] According to an embodiment of the present invention, model parameters of an initial encoder to be trained and an ECG decoder to be trained are adjusted according to a modality conversion loss value to obtain a trained encoder and ECG decoder.

[0157] According to an embodiment of the present invention, parameters of an electromagnetic encoder to be trained, an ECG encoder to be trained and an ECG decoder to be trained can be adjusted based on the modal conversion loss value, so that the mechanical motion state characteristics generated by the electromagnetic encoder to be trained can have a high feature similarity with the ECG motion state characteristics generated by the ECG encoder to be trained, and the ultrasonic ECG reconstructed signal reconstructed and restored by the ECG decoder to be trained can be consistent with the original ultrasonic ECG signal sample, thereby obtaining a trained encoder and ECG decoder.

[0158] According to an embodiment of the present invention, a trained encoder is determined as an encoder of the electromechanical modal conversion model to be trained.

[0159] According to an embodiment of the present invention, the ECG decoder can be used only to assist in training the encoder to be trained, so that only the electromagnetic encoder and ECG encoder obtained after training need to be retained as trained encoders, so as to utilize the trained encoder to assist in the training of the ultrasound cardiography video sub-model to be trained.

[0160] According to an embodiment of the present invention, an initial encoder to be trained and an ECG decoder to be trained are first obtained, and electromagnetic echo signal samples and ultrasonic ECG signal samples in a training sample set are respectively input into the electromagnetic encoder to be trained and the ECG encoder to be trained for training and learning, and mechanical motion state feature samples and ECG motion state feature samples are respectively obtained, and then the ECG decoder to be trained is trained using the ECG motion state feature samples output by the ECG encoder to be trained, so that the ECG decoder to be trained can reconstruct an ultrasonic ECG reconstruction signal, and based on a predetermined modal conversion loss function, a modal conversion loss value is obtained according to the mechanical motion state feature samples, the ECG motion state feature samples, the ultrasonic ECG reconstruction signal and the ultrasonic ECG signal samples, and the modal conversion loss value to be trained is adjusted. The model parameters of the initial encoder and the ECG decoder to be trained are obtained to obtain the trained encoder and ECG decoder. Since the subsequent training of the ultrasound video sub-model only requires the trained electromagnetic encoder and ECG encoder, the trained electromagnetic encoder and ECG encoder are retained as the encoder of the electromechanical modal conversion model, and the pre-training of the encoder is realized. Through repeated training, the mechanical motion state feature samples output by the encoder can be infinitely close to the ECG motion state feature samples, and infinitely close to the real cardiac mechanical activity features, so that when the ECG motion state feature samples are used to train the ultrasound video sub-model to be trained, it can be equivalent to using the mechanical motion state feature samples to train the ultrasound video sub-model to be trained. Further, in addition to adjusting the parameters of the electromagnetic encoder and ECG encoder to be trained according to the similarity between the mechanical motion state feature samples and the ECG motion state feature samples, the parameters of the electromagnetic encoder and ECG encoder to be trained can also be adjusted according to the ultrasound ECG reconstruction signal restored by the ECG decoder to be trained, while improving the accuracy of the encoder output, the correlation between the encoder, decoder and ultrasound video sub-model to be trained can also be improved.

[0161] According to an embodiment of the present invention, when the trained electromechanical modal conversion model is used to process the target electromagnetic echo signal extracted from actual measurement, the encoder only needs to use the trained electromagnetic encoder to perform feature mapping processing on the target electromagnetic echo signal.

[0162] Figure 3a A schematic diagram showing an actual echocardiogram according to an embodiment of the present invention is shown. Figure 3b A schematic diagram of a radar image of a heart obtained by using the method of the present invention according to an embodiment of the present invention is shown.

[0163] like Figure 3a~3b As shown, Figure 3a~3bThe radar image and the ultrasound cardiogram of the heart obtained by the method of the present invention are shown. By comparison, it can be seen that the structure shape and the heat map distribution between the radar image obtained by the method of the present invention and the actual ultrasound cardiogram are basically consistent.

[0164] Figure 4 A schematic diagram showing a comparison between a partial mechanical motion state characteristic obtained according to an embodiment of the present invention and a left ventricular volume obtained from an actual partial cardiac motion state is shown.

[0165] like Figure 4 As shown, Figure 4 The figure shows a comparison between the left ventricular volume obtained by the partial mechanical motion state characteristics and the actual partial cardiac motion state. The solid line can be represented as the left ventricular volume obtained by the actual partial cardiac motion state, and the dotted line can be represented as the left ventricular volume obtained by the partial mechanical motion state characteristics obtained according to the present invention. The minimum point A of the curve can be represented as the cardiac contraction frame, the maximum point B of the curve can be represented as the cardiac diastole frame, 401 can be represented as the rapid filling period, 402 can be represented as the global diastole period, 403 can be represented as the atrial contraction period, 404 can be represented as the isovolumetric contraction period, and 405 can be represented as the ejection period. It can be seen from the figure that the change trend of the left ventricular volume obtained by the method of the present invention is consistent with the actual change trend.

[0166] Figure 5 A schematic diagram showing a comparison between the total mechanical motion state characteristics obtained according to an embodiment of the present invention and the left ventricular volume obtained from the actual total cardiac motion state is shown.

[0167] like Figure 5 As shown, Figure 5 The figure shows a comparison between the left ventricular volume obtained by the partial mechanical motion state characteristics and the actual partial cardiac motion state. The solid line can be represented as the left ventricular volume obtained by the actual partial cardiac motion state, and the dotted line can be represented as the left ventricular volume obtained by the partial mechanical motion state characteristics obtained according to the present invention. It can be seen from the figure that the change trend of the left ventricular volume obtained by the method of the present invention is consistent with the actual change trend.

[0168] Figure 6 A structural block diagram of a device for generating a cardiac video based on an electromagnetic signal according to an embodiment of the present invention is shown.

[0169] like Figure 6 As shown, the device for generating a cardiac video based on an electromagnetic signal in this embodiment includes: a determination module 610 , a feature mapping module 620 and a generation module 630 .

[0170] The determination module 610 is used to determine the target electromagnetic echo signal corresponding to the heart region of the object to be measured, wherein the target electromagnetic echo signal includes the heartbeat motion feature after eliminating the respiratory motion feature. The determination module 610 can be used to perform the operation S210 described above, which will not be described in detail here.

[0171] The feature mapping module 620 is used to perform feature mapping processing on the target electromagnetic echo signal using the encoder of the electromechanical modal conversion model to obtain the target mechanical motion state characteristics of the heart area, wherein the target mechanical motion state characteristics correspond to the characteristics of the electrocardiographic motion state of the object to be measured. The feature mapping module 620 can be used to perform the operation S220 described above, which will not be repeated here.

[0172] The generating module 630 is used to process the target mechanical motion state characteristics and the conditional image using the ultrasound video sub-model of the electromechanical modal conversion model to generate a target ultrasound video of the object to be tested, wherein the conditional image represents the ultrasound template image. The generating module 630 can be used to perform the operation S230 described above, which will not be described in detail here.

[0173] According to an embodiment of the present invention, the determination module 610 includes: a first acquisition submodule, a first determination submodule, a second determination submodule and a first processing submodule.

[0174] The first acquisition submodule is used to acquire a plurality of initial electromagnetic echo signals of the object to be measured, wherein the initial electromagnetic echo signals include a plurality of signals that are spatially overlapped with each other and reflected by a plurality of body regions of the object to be measured within a predetermined period of time.

[0175] The first determination submodule is used to perform fast Fourier transform on multiple initial electromagnetic echo signals to determine the two-way propagation distance of each initial electromagnetic echo signal, wherein the two-way propagation distance represents the propagation distance of the transmitted signal from the transmitting antenna to the body area and the propagation distance of the initial electromagnetic echo signal reflected from the body area to the receiving antenna.

[0176] The second determination submodule is used to determine a target two-way propagation distance from a plurality of two-way propagation distances based on the position information of the object to be measured.

[0177] The first processing submodule is used to process each initial electromagnetic echo signal corresponding to the target two-way propagation distance to obtain a target electromagnetic echo signal.

[0178] According to an embodiment of the present invention, the first processing submodule includes: a first determining unit, a second determining unit, a third determining unit and a first filtering unit.

[0179] The first determining unit is used to determine each initial electromagnetic echo signal corresponding to the target round-trip propagation distance into a plurality of intermediate electromagnetic echo signals.

[0180] The second determination unit is used to perform fast Fourier transform on the multiple intermediate electromagnetic echo signals to determine the propagation angle of each intermediate electromagnetic echo signal, wherein the propagation angle is characterized by the angle between the intermediate electromagnetic echo signal and the object to be measured.

[0181] The third determining unit is used to determine a target propagation angle from multiple propagation angles based on angle information of the object to be measured relative to the receiving antenna.

[0182] The first filtering unit is used to perform second-order differential filtering on the intermediate electromagnetic echo signal corresponding to the target propagation angle to obtain the target electromagnetic echo signal.

[0183] According to an embodiment of the present invention, the generation module 630 includes: a second acquisition submodule and a first generation submodule.

[0184] The second acquisition submodule is used to acquire a Gaussian noise matrix for the ultrasound cardiography video.

[0185] The first generation submodule is used to input the Gaussian noise matrix, the target mechanical motion state characteristics and the conditional image into the ultrasonic video submodel. Based on the target mechanical motion state characteristics and the conditional image, the ultrasonic video submodel denoises the Gaussian noise matrix to generate the target ultrasonic video of the object to be measured.

[0186] According to an embodiment of the present invention, the device for generating a cardiac video based on an electromagnetic signal further includes: a training module.

[0187] The training module is used to train the electromechanical modal conversion model to be trained.

[0188] According to an embodiment of the present invention, the training module includes: a third acquisition submodule, a first mapping submodule, a second generation submodule, a first obtaining submodule and a first adjustment submodule.

[0189] The third acquisition submodule is used to acquire the electromechanical modal conversion model to be trained and the training sample data set, wherein the electromechanical modal conversion model to be trained includes an encoder and an ultrasonic video submodel to be trained, and the training sample data set includes electromagnetic echo signal samples, ultrasonic electrocardiogram signal samples, ultrasonic video samples and Gaussian noise samples.

[0190] The first mapping submodule is used to input the ultrasonic electrocardiogram signal sample into the encoder for feature mapping processing, and output the electrocardiogram motion state feature sample.

[0191] The second generation submodule is used to input the electrocardiogram motion state feature samples and the ultrasonic cardiogram video samples into the ultrasonic cardiogram video submodel to be trained for processing, so as to generate training ultrasonic video noise features.

[0192] The first obtaining submodule is used to obtain a video noise loss value based on a video noise loss function and according to training ultrasound video noise features and Gaussian noise samples.

[0193] The first adjustment submodule is used to adjust the model parameters of the ultrasound cardiology video submodel to be trained according to the video noise loss value to obtain the trained electromechanical mode conversion model.

[0194] According to an embodiment of the present invention, the first acquisition submodule includes: a first acquisition unit, a first mapping unit, a first obtaining unit, a first reconstruction unit, a second obtaining unit, a first adjustment unit and a fourth determination unit.

[0195] The first acquisition unit is used to acquire an initial encoder to be trained and an ECG decoder to be trained, wherein the initial encoder includes an electromagnetic encoder to be trained and an ECG encoder to be trained.

[0196] The first mapping unit is used to input the electromagnetic echo signal sample to the electromagnetic encoder to be trained for feature mapping processing to generate a mechanical motion state feature sample.

[0197] The first obtaining unit is used to input the ultrasonic electrocardiogram signal sample into the electrocardiogram encoder to be trained for feature mapping processing to obtain the electrocardiogram motion state feature sample.

[0198] The first reconstruction unit is used to input the ECG motion state feature samples into the ECG decoder to be trained for signal reconstruction processing to obtain an ultrasonic ECG reconstruction signal.

[0199] The second obtaining unit is used to obtain a modal conversion loss value based on a predetermined modal conversion loss function, according to the mechanical motion state feature samples, the electrocardiographic motion state feature samples, the ultrasonic electrocardiographic reconstruction signal and the ultrasonic electrocardiographic signal samples.

[0200] The first adjustment unit is used to adjust the model parameters of the initial encoder to be trained and the ECG decoder to be trained according to the modal conversion loss value to obtain the trained encoder and ECG decoder.

[0201] The fourth determining unit is used to determine the trained encoder as the encoder of the electromechanical modal conversion model to be trained.

[0202] According to an embodiment of the present invention, the second obtaining unit includes: a fifth obtaining sub-unit, a sixth obtaining sub-unit and a seventh obtaining sub-unit.

[0203] The fifth obtaining subunit is used to obtain the motion state feature similarity according to a predetermined motion state feature similarity function, the mechanical motion state feature sample and the electrocardiogram motion state feature sample.

[0204] The sixth obtaining subunit is used to obtain the reconstructed signal feature similarity according to the predetermined reconstructed signal feature similarity function, the ultrasonic electrocardiogram reconstructed signal and the ultrasonic electrocardiogram signal sample.

[0205] The seventh obtaining subunit is used to obtain the modal conversion loss value according to a predetermined modal conversion loss function, a balance parameter, a motion state feature similarity and a reconstructed signal feature similarity.

[0206] According to an embodiment of the present invention, the second generating submodule includes: a fourth obtaining unit and a first denoising unit.

[0207] The fourth obtaining unit is used to add Gaussian noise samples to the ultrasonic cardiology video samples to obtain ultrasonic cardiology noise video.

[0208] The first denoising unit is used to input the ultrasonic cardiology noise video, the electrocardiogram motion state characteristic sample and the ultrasonic cardiology video sample into the ultrasonic cardiology video sub-model to be trained for denoising, and output the training ultrasonic video noise characteristic.

[0209] According to an embodiment of the present invention, any multiple modules of the determination module 610, the feature mapping module 620 and the generation module 630 can be combined into one module for implementation, or any one of the modules can be split into multiple modules. Alternatively, at least part of the functions of one or more of these modules can be combined with at least part of the functions of other modules and implemented in one module. According to an embodiment of the present invention, at least one of the determination module 610, the feature mapping module 620 and the generation module 630 can be at least partially implemented as a hardware circuit, such as a field programmable gate array (FPGA), a programmable logic array (PLA), a system on a chip, a system on a substrate, a system on a package, an application specific integrated circuit (ASIC), or can be implemented by hardware or firmware such as any other reasonable way of integrating or packaging the circuit, or implemented in any one of the three implementation methods of software, hardware and firmware or in any appropriate combination of any of them. Alternatively, at least one of the determination module 610, the feature mapping module 620 and the generation module 630 can be at least partially implemented as a computer program module, and when the computer program module is run, the corresponding function can be executed.

[0210] Figure 7 A block diagram of an electronic device for a method of generating a cardiac video based on an electromagnetic signal according to an embodiment of the present invention is shown.

[0211] like Figure 7As shown, the electronic device according to an embodiment of the present invention includes a processor 701, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 702 or a program loaded from a storage part 708 to a random access memory (RAM) 703. The processor 701 may include, for example, a general-purpose microprocessor (e.g., a CPU), an instruction set processor and / or a related chipset and / or a special-purpose microprocessor (e.g., an application-specific integrated circuit (ASIC)), etc. The processor 701 may also include an onboard memory for caching purposes. The processor 701 may include a single processing unit or multiple processing units for performing different actions of the method flow according to an embodiment of the present invention.

[0212] In RAM 703, various programs and data required for the operation of the electronic device are stored. The processor 701, ROM 702 and RAM 703 are connected to each other via a bus 704. The processor 701 performs various operations of the method flow according to the embodiment of the present invention by executing the programs in ROM 702 and / or RAM 703. It should be noted that the program can also be stored in one or more memories other than ROM 702 and RAM 703. The processor 701 can also perform various operations of the method flow according to the embodiment of the present invention by executing the programs stored in the one or more memories.

[0213] According to an embodiment of the present invention, the electronic device may further include an input / output (I / O) interface 705, which is also connected to the bus 704. The electronic device may further include one or more of the following components connected to the I / O interface 705: an input portion 706 including a keyboard, a mouse, etc.; an output portion 707 including a cathode ray tube (CRT), a liquid crystal display (LCD), etc., and a speaker, etc.; a storage portion 708 including a hard disk, etc.; and a communication portion 709 including a network interface card such as a LAN card, a modem, etc. The communication portion 709 performs communication processing via a network such as the Internet. A drive 710 is also connected to the I / O interface 705 as needed. A removable medium 711, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., is installed on the drive 710 as needed, so that a computer program read therefrom is installed into the storage portion 708 as needed.

[0214] The present invention also provides a computer-readable storage medium, which may be included in the device / apparatus / system described in the above embodiment; or may exist independently without being assembled into the device / apparatus / system. The above computer-readable storage medium carries one or more programs, and when the above one or more programs are executed, the method according to the embodiment of the present invention is implemented.

[0215] According to an embodiment of the present invention, the computer-readable storage medium may be a non-volatile computer-readable storage medium, for example, it may include but is not limited to: a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof. In the present invention, the computer-readable storage medium may be any tangible medium containing or storing a program, which may be used by or in combination with an instruction execution system, an apparatus or a device. For example, according to an embodiment of the present invention, the computer-readable storage medium may include the ROM 702 and / or RAM 703 described above and / or one or more memories other than ROM 702 and RAM 703.

[0216] The embodiment of the present invention also includes a computer program product, which includes a computer program, and the computer program contains program code for executing the method shown in the flowchart. When the computer program product is run in a computer system, the program code is used to enable the computer system to implement the method for generating a cardiac video based on an electromagnetic signal provided by the embodiment of the present invention.

[0217] The computer program executes the above functions defined in the system / device of the embodiment of the present invention when it is executed by the processor 701. According to the embodiment of the present invention, the system, device, module, unit, etc. described above can be implemented by a computer program module.

[0218] In one embodiment, the computer program may rely on tangible storage media such as optical storage devices, magnetic storage devices, etc. In another embodiment, the computer program may also be transmitted and distributed in the form of signals on a network medium, and downloaded and installed through the communication part 709, and / or installed from the removable medium 711. The program code contained in the computer program may be transmitted using any appropriate network medium, including but not limited to: wireless, wired, etc., or any suitable combination of the above.

[0219] In such an embodiment, the computer program can be downloaded and installed from the network through the communication part 709, and / or installed from the removable medium 711. When the computer program is executed by the processor 701, the above functions defined in the system of the embodiment of the present invention are performed. According to the embodiment of the present invention, the system, device, means, module, unit, etc. described above can be implemented by a computer program module.

[0220] It will be appreciated by those skilled in the art that the features described in the various embodiments of the present invention may be combined and / or combined in various ways, even if such combinations and / or combinations are not explicitly described in the present invention. In particular, without departing from the spirit and teachings of the present invention, the features described in the various embodiments of the present invention may be combined and / or combined in various ways. All of these combinations and / or combinations fall within the scope of the present invention.

Claims

1. A method for generating a cardiac video based on an electromagnetic signal, characterized in that: include: Determining a target electromagnetic echo signal corresponding to a heart region of the subject to be measured, wherein the target electromagnetic echo signal includes a heartbeat motion feature after eliminating a respiratory motion feature; Performing feature mapping processing on the target electromagnetic echo signal using an encoder of an electromechanical modal conversion model to obtain target mechanical motion state characteristics of the heart region, wherein the target mechanical motion state characteristics correspond to the characteristics of the electrocardiographic motion state of the object to be measured; The target mechanical motion state characteristics and conditional image are processed using the ultrasonic cardiography video sub-model of the electromechanical modal conversion model to generate a target ultrasonic cardiography video of the object to be measured, wherein the conditional image represents an ultrasonic cardiography template image.

2. The method according to claim 1, characterized in that The step of determining a target electromagnetic echo signal corresponding to a heart region of the object to be measured comprises: Acquire a plurality of initial electromagnetic echo signals of the object to be measured, wherein the initial electromagnetic echo signals include a plurality of signals spatially overlapped with each other and reflected by a plurality of body regions of the object to be measured within a predetermined period of time; Performing fast Fourier transform on the multiple initial electromagnetic echo signals to determine a two-way propagation distance of each of the initial electromagnetic echo signals, wherein the two-way propagation distance represents a propagation distance of a transmission signal transmitted from a transmitting antenna to the body region and a propagation distance of the initial electromagnetic echo signal reflected from the body region to a receiving antenna; Determining a target two-way propagation distance from a plurality of the two-way propagation distances based on the position information of the object to be measured; Each of the initial electromagnetic echo signals corresponding to the target round-trip propagation distance is processed to obtain the target electromagnetic echo signal.

3. The method according to claim 2, characterized in that The processing of each of the initial electromagnetic echo signals corresponding to the target two-way propagation distance to obtain the target electromagnetic echo signal includes: Determining each of the initial electromagnetic echo signals corresponding to the target round-trip propagation distance as a plurality of intermediate electromagnetic echo signals; Performing a fast Fourier transform on the multiple intermediate electromagnetic echo signals to determine a propagation angle of each of the intermediate electromagnetic echo signals, wherein the propagation angle is characterized by an angle between the intermediate electromagnetic echo signal and the object to be measured; Determining a target propagation angle from a plurality of propagation angles based on the angle information of the object to be measured relative to the receiving antenna; A second-order differential filtering process is performed on the intermediate electromagnetic echo signal corresponding to the target propagation angle to obtain the target electromagnetic echo signal.

4. The method according to claim 1, characterized in that: The method of processing the target mechanical motion state characteristics and conditional images using the ultrasonic cardiography video sub-model of the electromechanical modal conversion model to generate the target ultrasonic cardiography video of the object to be measured includes: Obtaining a Gaussian noise matrix for echocardiography video; The Gaussian noise matrix, the target mechanical motion state characteristics and the conditional image are input into the ultrasonic video sub-model. Based on the target mechanical motion state characteristics and the conditional image, the ultrasonic video sub-model denoises the Gaussian noise matrix to generate the target ultrasonic video of the object to be measured.

5. The method according to claim 1, characterized in that The electromechanical mode conversion model is trained using the following methods, including: Acquire an electromechanical modal conversion model to be trained and a training sample data set, wherein the electromechanical modal conversion model to be trained includes an encoder and an ultrasonic video sub-model to be trained, and the training sample data set includes electromagnetic echo signal samples, ultrasonic electrocardiogram signal samples, ultrasonic video samples and Gaussian noise samples; Inputting the ultrasonic electrocardiogram signal sample into the encoder for feature mapping processing to generate electrocardiogram motion state feature samples; Inputting the electrocardiographic motion state feature samples and the ultrasonic cardiography video samples into the ultrasonic cardiography video sub-model to be trained for processing, and outputting training ultrasonic video noise features; Based on the video noise loss function, a video noise loss value is obtained according to the training ultrasound video noise feature and the Gaussian noise sample; According to the video noise loss value, the model parameters of the ultrasonic cardiology video sub-model to be trained are adjusted to obtain a trained electromechanical mode conversion model.

6. The method according to claim 5, characterized in that The encoder of the electromechanical modal conversion model to be trained is trained in the following manner, including: Acquire an initial encoder to be trained and an ECG decoder to be trained, wherein the initial encoder includes an electromagnetic encoder to be trained and an ECG encoder to be trained; Inputting the electromagnetic echo signal sample into the electromagnetic encoder to be trained for feature mapping processing to generate a mechanical motion state feature sample; Inputting the ultrasonic electrocardiogram signal sample into the electrocardiogram encoder to be trained for feature mapping processing to obtain an electrocardiogram motion state feature sample; Inputting the ECG motion state characteristic sample into the ECG decoder to be trained for signal reconstruction processing to obtain an ultrasonic ECG reconstruction signal; Obtaining a modal conversion loss value according to a predetermined modal conversion loss function, the mechanical motion state characteristic sample, the electrocardiographic motion state characteristic sample, the ultrasonic electrocardiographic reconstruction signal and the ultrasonic electrocardiographic signal sample; According to the modal conversion loss value, adjusting the model parameters of the initial encoder to be trained and the ECG decoder to be trained to obtain a trained encoder and ECG decoder; The trained encoder is determined as the encoder of the electromechanical modal conversion model to be trained.

7. The method according to claim 6, characterized in that The method of obtaining a modal conversion loss value according to a predetermined modal conversion loss function, the mechanical motion state characteristic sample, the electrocardiographic motion state characteristic sample, the ultrasonic electrocardiographic reconstruction signal and the ultrasonic electrocardiographic signal sample comprises: Obtaining motion state feature similarity according to a predetermined motion state feature similarity function, the mechanical motion state feature sample and the electrocardiogram motion state feature sample; Obtaining a reconstructed signal feature similarity according to a predetermined reconstructed signal feature similarity function, the ultrasonic electrocardiogram reconstructed signal and the ultrasonic electrocardiogram signal sample; The modal conversion loss value is obtained according to a predetermined modal conversion loss function, a balance parameter, the motion state feature similarity and the reconstructed signal feature similarity.

8. The method according to claim 5, characterized in that The step of inputting the electrocardiographic motion state feature sample and the ultrasonic cardiography video sample into the ultrasonic cardiography video sub-model to be trained for processing and outputting the training ultrasonic cardiography video noise feature comprises: Adding the Gaussian noise sample to the ultrasonic cardiography video sample to obtain an ultrasonic cardiography noise video; The ultrasonic cardiology noise video, the electrocardiographic motion state characteristic sample and the ultrasonic cardiology video sample are input into the ultrasonic cardiology video sub-model to be trained for denoising, and the training ultrasonic cardiology video noise characteristic is output.

9. A device for generating a cardiac video based on an electromagnetic signal, characterized in that: The device comprises: A determination module, used to determine a target electromagnetic echo signal corresponding to a heart region of the object to be measured, wherein the target electromagnetic echo signal includes a heartbeat motion feature after eliminating a respiratory motion feature; a feature mapping module, configured to perform feature mapping processing on the target electromagnetic echo signal using an encoder of an electromechanical modal conversion model to obtain a target mechanical motion state feature of the heart region, wherein the target mechanical motion state feature corresponds to a feature of the electrocardiographic motion state of the subject to be measured; A generation module is used to process the target mechanical motion state characteristics and conditional images using the ultrasonic video sub-model of the electromechanical modal conversion model to generate a target ultrasonic video of the object to be measured, wherein the conditional image represents an ultrasonic template image.

10. An electronic device, characterized in that: include: one or more processors; a storage device for storing one or more programs, When the one or more programs are executed by the one or more processors, the one or more processors execute the method according to any one of claims 1 to 8.

Images