Cardiac compression monitoring method, apparatus, device, and medium

Abstract

This invention relates to the field of cardiac monitoring technology and discloses a cardiac compression monitoring method: After extracorporeal membrane oxygenation (ECMO) is established, the target object is monitored in real time using ultrasound equipment to obtain an initial cardiac image; image analysis results are acquired and input into a compression device, which then performs compression operations based on the analysis results; if the interval reaches a preset period, an updated cardiac image of the target object is acquired using ultrasound equipment; updated analysis results are acquired and input into the compression device, which then performs compression operations based on the updated analysis results. This invention utilizes ultrasound equipment to achieve real-time monitoring of the target object and acquisition of cardiac images, thereby enabling the detection of thrombi in the cardiac images. The compression device accelerates blood flow, thus reducing the risk of thrombus formation and the workload of medical personnel.

Description

Technical Field

[0001] This invention relates to the field of cardiac monitoring technology, and in particular to a method, device, equipment and medium for cardiac compression monitoring. Background Technology

[0002] ECMO (Extracorporeal Membrane Oxygenation) is a cardiopulmonary bypass technique used in cardiac surgery. The principle is to draw venous blood out of the body, oxygenate it through a special artificial heart-lung bypass, and then inject it into the patient's arterial or venous system, thus partially replacing the heart and lungs and maintaining the oxygenated blood supply to the body's organs and tissues.

[0003] However, because extracorporeal membrane oxygenation (ECMO) replaces the functions of the heart and lungs, and the blood pump replaces the heart's contraction and relaxation, blood from the heart cannot effectively participate in the overall circulatory system, resulting in a large residual blood volume and slow blood flow within the heart. This makes patients prone to pulmonary embolism during ECMO treatment. Blood clots reduce the oxygenation efficiency of the membrane oxygenator, requiring clinicians to perform repeated chest compressions during ECMO to ensure blood flow to the heart and prevent clot formation. Summary of the Invention

[0004] This invention provides a method, device, equipment, and medium for monitoring cardiac compression, in order to solve the problem in the prior art that clinicians need to frequently compress the heart during ECMO treatment to prevent thrombosis.

[0005] A method for monitoring cardiac compression, comprising:

[0006] After extracorporeal membrane oxygenation (ECMO) is established, the target object corresponding to the establishment of ECMO is monitored in real time using ultrasound equipment to obtain the initial cardiac image corresponding to the target object.

[0007] Obtain the image analysis result corresponding to the initial heart image, and input the image analysis result into the compression device, so that the compression device performs a compression operation on the target object according to the image analysis result;

[0008] If the interval reaches the preset interval period, an updated cardiac image of the target object is acquired by ultrasound equipment;

[0009] Obtain the update analysis result corresponding to the updated heart image, and input the update analysis result into the compression device, so that the compression device performs a compression operation on the target object according to the update analysis result.

[0010] A cardiac compression monitoring device, comprising:

[0011] The initial image acquisition module is used to monitor the target object corresponding to the establishment of extracorporeal membrane oxygenation (ECMO) in real time using ultrasound equipment after the establishment of ECMO, and obtain the initial cardiac image corresponding to the target object.

[0012] The initial compression operation module is used to acquire the image analysis result corresponding to the initial heart image, and input the image analysis result into the compression device, so that the compression device performs a compression operation on the target object according to the image analysis result;

[0013] An updated image acquisition module is used to acquire updated cardiac images of the target object via ultrasound equipment if the interval time reaches a preset interval period.

[0014] The update compression operation module is used to obtain the update analysis result corresponding to the updated heart image, and input the update analysis result into the compression device, so that the compression device performs a compression operation on the target object according to the update analysis result.

[0015] A computer device includes a memory, a controller, and a computer program stored in the memory and executable on the controller, wherein the controller executes the computer program to implement the aforementioned cardiac compression monitoring method.

[0016] A computer-readable storage medium storing a computer program that, when executed by a controller, implements the aforementioned cardiac compression monitoring method.

[0017] The cardiac compression monitoring method, device, equipment, and medium provided by this invention achieve real-time monitoring of the target object through ultrasound equipment, thereby enabling the acquisition of cardiac images. Image analysis results allow for the detection of the risk of thrombosis in the cardiac images, thus achieving precise control of the compression device. The compression device accelerates blood flow to the heart, reducing the risk of thrombosis and facilitating the replacement of medical personnel, thereby reducing their workload. Attached Figure Description

[0018] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments of the present invention will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 This is a flowchart of a cardiac compression monitoring method according to an embodiment of the present invention;

[0020] Figure 2 This is a schematic diagram of a cardiac compression monitoring device according to an embodiment of the present invention. Detailed Implementation

[0021] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0022] This invention provides a method for monitoring cardiac compression, in one embodiment, such as... Figure 1 As shown, its technical solution mainly includes the following steps:

[0023] S10, after the establishment of extracorporeal membrane oxygenation (ECMO), the target object corresponding to the establishment of ECMO is monitored in real time using ultrasound equipment to obtain an initial cardiac image corresponding to the target object.

[0024] In essence, extracorporeal membrane oxygenation (ECMO) involves drawing venous blood from the body, oxygenating it through a specially designed artificial heart-lung bypass, and then injecting it back into the patient's arterial or venous system, thus partially replacing the heart and lungs and maintaining oxygenated blood supply to the body's organs and tissues. Ultrasound equipment refers to a series of instruments that operate based on the principles of ultrasound. Initial cardiac images refer to images of the target object's heart acquired using ultrasound equipment.

[0025] Specifically, extracorporeal membrane oxygenation (ECMO) is established on the target subject, which means using an extracorporeal membrane oxygenation system to replace the functions of the heart and lungs, and a blood pump to replace the heart's contraction and relaxation. After ECMO is established, the target subject corresponding to the establishment of ECMO is monitored in real time using ultrasound equipment. That is, preset locations are obtained, and an ultrasound probe is placed at the preset location to detect that location. Images of each preset location are acquired based on the ultrasound waves, thereby obtaining initial cardiac images of the target subject containing multiple locations.

[0026] S20, obtain the image analysis result corresponding to the initial heart image, and input the image analysis result into the compression device, so that the compression device performs a compression operation on the target object according to the image analysis result.

[0027] Understandably, image analysis results refer to the results of analysis of initial cardiac images to characterize the presence of a risk of blood clot formation. A compression device refers to a device used to apply pressure to the heart portion of a target object.

[0028] Specifically, an image recognition model is acquired, and an initial heart image is input into the model. The model, learned during training, performs risk assessment on cross-sectional images at different locations within the initial heart image, outputting the risk of thrombosis formation and its probability. This risk and probability are then used as the corresponding image analysis result for the initial heart image. A risk threshold is set: if the identified probability exceeds the threshold, a risk of thrombosis is identified, and the image analysis result representing the probability of thrombosis is output; if the probability does not exceed the threshold, no risk of thrombosis is identified, and the image analysis result representing the probability of thrombosis is output. The image analysis result is then input into a compression device. By recognizing the image analysis result, the presence of a risk of thrombosis is determined. If no risk of thrombosis is found, compression is not performed on the heart. If a risk of thrombosis is found, the probability of thrombosis is obtained, and the corresponding compression depth, frequency, and duration are determined. The compression device is then controlled to perform compressions at these specified depths, frequencies, and durations.

[0029] S30, if the interval time reaches the preset interval period, then the updated cardiac image of the target object is acquired by ultrasound equipment.

[0030] Understandably, a preset interval period refers to the time interval between two pre-set compressions, for example, 1 hour. The interval time refers to the time since the last compression. Updating the heart image refers to acquiring an image of the target subject's heart using an ultrasound device before a new compression.

[0031] Specifically, the system records the time of the last compression operation and calculates the interval between the current time and the last compression operation in real time. This interval is compared with a preset interval period. If the interval period reaches the preset interval period, the ultrasound probe in the ultrasound device is controlled to re-acquire images of the preset position, thereby obtaining an updated cardiac image corresponding to the target object.

[0032] S40, obtain the update analysis result corresponding to the updated heart image, and input the update analysis result into the compression device, so that the compression device performs a compression operation on the target object according to the update analysis result.

[0033] Understandably, updated analysis results refer to the results of analysis of updated cardiac images, used to characterize the risk of thrombosis.

[0034] Specifically, an image recognition model is acquired, and the updated heart image is input into the model. The model, learned during training, performs risk assessment on cross-sectional images at different locations within the updated heart image, outputting the risk of thrombosis formation and its probability. The presence and probability of thrombosis formation are then defined as the update analysis result corresponding to the updated heart image. Specifically, if the identified probability of thrombosis formation exceeds a risk threshold, an update analysis result indicating a risk of thrombosis formation is output; if the probability does not exceed the risk threshold, an update analysis result indicating no risk of thrombosis formation is output. This update analysis result is then input into a compression device. By recognizing the update analysis result, the device determines whether a risk of thrombosis formation exists. If no risk of thrombosis formation exists, the compression device is not used to perform compression on the target's heart. If there is a risk of thrombosis, the probability of thrombosis formation is obtained, and the corresponding compression depth, compression frequency, and compression time are retrieved based on this probability. The compression device is then controlled to perform compression operations at the specified compression depth, frequency, and duration. In another embodiment, the updated cardiac image is transmitted to a preset client, and the updated analysis results are received from the preset client.

[0035] This invention utilizes ultrasound equipment to achieve real-time monitoring of the target object, thereby enabling the acquisition of cardiac images. Image analysis results allow for the detection of the risk of thrombosis in the cardiac images, thus enabling precise control of the compression device. The compression device accelerates blood flow to the heart, reducing the risk of thrombosis and facilitating the replacement of medical personnel, thereby reducing their workload.

[0036] In one embodiment, step S10, which involves real-time monitoring of the target object corresponding to the establishment of extracorporeal membrane oxygenation using ultrasound equipment to obtain an initial cardiac image corresponding to the target object, includes:

[0037] S101, control the ultrasound probe in the ultrasound device to detect each preset position, so as to acquire images of different cardiac sections of the target object and obtain an initial cardiac image containing multiple section images.

[0038] Understandably, an ultrasound probe (also known as an ultrasound transducer) is a device that converts electrical energy into ultrasonic energy and transmits it into the human body, while simultaneously converting the ultrasonic signals reflected back from the body into electrical signals. A preset position refers to a pre-set detection location, such as the 3rd or 4th intercostal space at the left sternal border.

[0039] Specifically, at least one preset position is obtained, and the ultrasound probe in the ultrasound device is controlled to detect each preset position. That is, the ultrasound probe placed at the preset position collects information about the position and acquires images of the heart condition at each preset position, thereby obtaining cross-sectional images corresponding to each preset position, and determining all cross-sectional images as the initial heart image.

[0040] In one specific embodiment, when the preset position is a parasternal long-axis view: the ultrasound probe is placed at the left sternal border between the 3rd and 4th ribs, and the probe angle and direction are adjusted so that the ultrasound beam is parallel to the long axis of the heart. This view allows observation of structures such as the left atrium, left ventricle, mitral valve, aortic valve, and interventricular septum. When the preset position is a parasternal short-axis view: based on the parasternal long-axis view, the ultrasound probe is rotated approximately 90 degrees clockwise to obtain short-axis views at different levels, including short-axis views at the mitral valve level, papillary muscle level, and apical level. This view allows observation of the wall motion of each segment of the left ventricle, the morphology and size of the right ventricle, and structures such as the pulmonary valve. When the preset position is an apical four-chamber view: the ultrasound probe is placed at the apical pulsation point, pointing towards the right sternoclavicular joint, to obtain an apical four-chamber view. This view displays the four chambers of the heart (left atrium, left ventricle, right atrium, and right ventricle) and the mitral and tricuspid valves. When the preset position is the apical two-chamber cardiac view: Based on the apical four-chamber cardiac view, rotate the ultrasound probe slightly (usually about 30 degrees counterclockwise) to obtain the apical two-chamber cardiac view, mainly to observe the left atrium, left ventricle and mitral valve structure.

[0041] In this embodiment, by moving the ultrasound probe to different positions, images of the heart from different sections were acquired, thereby ensuring the accuracy of subsequent image recognition and the precise control of subsequent compression operations.

[0042] In one embodiment, step S20, namely obtaining the image parsing result corresponding to the initial cardiac image, includes:

[0043] S201, Obtain the image recognition model and input the initial heart image into the image recognition model.

[0044] S202, the initial heart image is subjected to image recognition using the image recognition model to obtain at least one cross-sectional image.

[0045] S203, thrombosis identification is performed on all the cross-sectional images using the image recognition model to obtain the image recognition result corresponding to each cross-sectional image.

[0046] S204, Based on all the image recognition results, determine the image parsing result corresponding to the initial heart image.

[0047] In essence, an image recognition model refers to a network structure used to identify the risk of thrombosis formation in cardiac images. This model is trained using a large amount of positive and negative sample data. A cross-sectional image refers to images acquired at different preset locations. The image recognition result is used to characterize the probability of thrombosis formation in that cross-sectional image. A thrombus is a solid mass formed in the heart and blood vessels of a living organism due to the coagulation of blood or the aggregation of certain formed elements in the blood.

[0048] Specifically, after obtaining the initial heart image, an image recognition model is acquired, and the initial heart image is input into the image recognition model. The image recognition model performs image recognition on the initial heart image, that is, it uses the image recognition capabilities learned during the training of the image recognition model to identify the cross-sectional images contained in the initial heart image, thereby obtaining at least one cross-sectional image. Then, the image recognition model performs thrombus identification on each cross-sectional image, that is, it uses the thrombus identification capabilities learned during the training of the image recognition model to identify the information in each cross-sectional image, thereby obtaining the image recognition result corresponding to each cross-sectional image. Next, based on all image recognition results, the image analysis result corresponding to the initial heart image is determined, that is, a preset risk threshold is obtained, and all image recognition results are compared one by one with the preset risk threshold. If all image recognition results do not exceed the preset risk threshold, an image analysis result representing the absence of thrombus formation and the probability of thrombus formation is obtained. If at least one image recognition result exceeds the preset risk threshold, an image analysis result representing the presence of thrombus formation and the probability of thrombus formation is obtained. When multiple image recognition results exceed the preset risk threshold, all formation probabilities are weighted to obtain the image analysis result.

[0049] In this embodiment, the initial heart image is identified by an image recognition model, which enables rapid and accurate identification of the initial heart image. This allows for the acquisition of image recognition results, which in turn enables the acquisition of image analysis results, facilitating the subsequent precise compression operation on the target object.

[0050] In one embodiment, step S203, which involves using the image recognition model to perform thrombus identification on all the cross-sectional images to obtain image recognition results corresponding to each cross-sectional image, includes:

[0051] S2031, feature extraction is performed on all the cross-sectional images through the convolutional layer in the image recognition model to obtain the convolutional features corresponding to each cross-sectional image.

[0052] S2032, All the convolutional features are pooled through the pooling layer in the image recognition model to obtain pooled features corresponding to each of the cross-sectional images.

[0053] S2033, Image recognition is performed on all the pooled features through the fully connected layer in the image recognition model to obtain the image recognition result corresponding to each of the cross-sectional images.

[0054] In essence, convolutional features refer to features extracted through a convolutional network. Pooling features refer to features processed by a pooling network.

[0055] Specifically, after obtaining the cross-sectional images, feature extraction is performed on all cross-sectional images through convolutional layers in the image recognition model. This involves traversing each cross-sectional image using a preset convolutional kernel and stride. The convolutional kernel slides across the cross-sectional image, extracting local features, and then processing these features using an activation function to obtain the convolutional features corresponding to each cross-sectional image. Next, pooling is performed on all convolutional features through pooling layers in the image recognition model. This involves traversing the entire convolutional feature map according to a preset pooling window size and stride. Starting from the top left corner of the convolutional feature map, the pooling window moves according to the stride, sequentially performing pooling on each region covered by the pooling window until all regions of the entire convolutional feature map have been traversed, thus obtaining the pooled features corresponding to each convolutional feature. If multiple convolutional layers exist, the image is processed through each layer before pooling. Alternatively, if multiple convolutional pooling layers exist, the image is processed one by one before proceeding to the next step. Next, image recognition is performed on all pooling features through the fully connected layer in the image recognition model. That is, the pooling features are calculated through each unit in the fully connected layer to obtain the image recognition result representing the probability of thrombus formation. In this way, the image recognition result corresponding to each cross-sectional image can be obtained.

[0056] In this embodiment, convolutional layers are used to extract features from each cross-sectional image, thus achieving the acquisition of convolutional features. Pooling layers are used to reduce the dimension of each convolutional feature, thus achieving the acquisition of pooling features, and consequently, the prediction of the probability of thrombosis formation, thereby achieving the acquisition of image recognition results.

[0057] In one embodiment, step S20, which involves instructing the pressing device to perform a pressing operation on the target object based on the image analysis result, includes:

[0058] S205, if the image analysis result indicates that the target object has a risk of thrombosis, then the probability of thrombosis is extracted from the image analysis result.

[0059] S206, Based on the formation probability, determine the pressing depth, pressing frequency, and pressing time corresponding to the target object.

[0060] S207, the pressing device is instructed to perform a pressing operation on the target object according to the pressing depth, the pressing frequency and the pressing time.

[0061] Understandably, the probability of formation refers to the likelihood of a blood clot forming. Compression depth, compression frequency, and compression time refer to the depth, frequency, and duration of compression corresponding to that probability. For example, the higher the probability, the deeper the compression, the higher the compression frequency, and the longer the compression time. The maximum compression depth cannot exceed 5 centimeters.

[0062] Specifically, after inputting the image analysis results into the compression device, if the image analysis results indicate a risk of thrombosis in the target object, the probability of thrombosis is extracted from the image analysis results. Then, based on the probability of thrombosis, the compression depth, compression frequency, and compression time corresponding to the target object are determined. That is, a preset mapping table is obtained, and the corresponding compression depth, compression frequency, and compression time are retrieved from the preset mapping table based on the probability of thrombosis and determined as the compression depth, compression frequency, and compression time corresponding to the target object. Next, the compression depth, compression frequency, and compression time are input into the compression device, and the compression device is instructed to perform compression operations on the target object according to the compression depth, compression frequency, and compression time. That is, the compression device is controlled to perform compression operations on the heart position of the target object at the specified compression depth and compression frequency, and the compression operation is stopped after the specified compression time.

[0063] In this embodiment, by generating probabilities, the compression depth, compression frequency, and compression time are obtained, thereby enabling precise control of the compression device and ensuring that the target object is not damaged during the compression process. Performing the compression operation through the compression device accelerates blood flow, thereby reducing the risk of thrombosis and the workload of medical staff.

[0064] In one embodiment, after step S20, i.e., after obtaining the image parsing result corresponding to the initial heart image, the method further includes:

[0065] S50, when the image analysis result indicates that the target object has a risk of thrombosis, a preset compression depth, a preset compression frequency, and a preset compression time are obtained, and the preset compression depth, the preset compression frequency, and the preset compression time are input into the compression device.

[0066] S60, the pressing device is instructed to perform a pressing operation on the target object according to the preset pressing depth, the preset pressing frequency and the preset pressing time.

[0067] Understandably, preset compression depth refers to the pre-set compression depth of the pressing device. Preset compression frequency refers to the pre-set number of compression operations per unit time. Preset compression time refers to the pre-set compression duration, for example, 5 minutes.

[0068] Specifically, after obtaining the image analysis result corresponding to the initial heart image, if the image analysis result indicates that the target object has a risk of thrombosis, a preset compression depth, preset compression frequency, and preset compression time are obtained. Then, the preset compression depth, preset compression frequency, and preset compression time are input into the compression device to control the device. The compression device is then instructed to perform compression operations on the target object according to the preset compression depth, preset compression frequency, and preset compression time; that is, the compression device is controlled to perform compression operations on the target object's heart position at the preset compression depth, preset compression frequency, and preset compression time to accelerate blood flow to that area.

[0069] In this embodiment, by preset pressing depth, preset pressing frequency, and preset pressing time, pressing of the target object is achieved, thereby avoiding changes to the pressing depth, pressing frequency, and pressing time, and thus facilitating the pressing operation of the target object through the pressing device.

[0070] In one embodiment, before step S201, i.e. before obtaining the image recognition model, the method further includes:

[0071] S701, Obtain a training dataset, wherein the training dataset includes at least one sample data and sample labels corresponding to each sample data.

[0072] In essence, sample data refers to cross-sectional images of different target objects showing both the presence and absence of thrombosis risk. Each sample data point is associated with a sample label, which characterizes whether the cross-sectional image of the sample data contains a thrombosis risk. Sample data and corresponding sample labels are obtained from different databases or clients, and a training dataset is constructed based on the obtained sample data and their corresponding sample labels.

[0073] S702, Obtain a preset training model, and perform image recognition on each of the sample data using the preset training model to obtain the sample recognition result.

[0074] Understandably, the pre-trained model refers to a pre-configured neural network used to identify the risk of thrombosis in cross-sectional images. This model's convolutional layers employ parameter sharing and sparse connections. Parameter sharing means that the same convolutional kernel shares the same weights at different locations in the image, allowing the model to learn the same features at different locations. Sparse connections mean that the convolutional kernel is connected to only a subset of pixels in the image, rather than to all pixels. The sample recognition result refers to the prediction result of the risk of thrombosis formation identified in the sample data.

[0075] Specifically, a pre-defined training model is obtained, and all sample data are input into it. The model preprocesses the sample data to produce images that meet pre-defined requirements. Then, convolutional layers in the model convolve the preprocessed sample data, with a pre-defined stride causing the convolution kernel to iterate through each preprocessed sample data point, resulting in convolutional features corresponding to each sample. Next, pooling layers in the model pool these features, with a pre-defined stride causing the pooling kernel to iterate through each feature, resulting in pooled features corresponding to each sample. This can be achieved by using multiple convolutional layers or multiple convolutional and pooling layers. Finally, a fully connected layer performs risk prediction on the pooled features corresponding to each sample, thus obtaining a sample identification result representing the risk of thrombosis.

[0076] S703, determine the prediction loss value of the preset training model based on the sample label and the sample recognition result corresponding to the same sample data.

[0077] Understandably, the predicted loss value is the loss value generated during the prediction process of the sample data.

[0078] Specifically, the sample recognition results corresponding to each sample data are arranged according to the order of the sample data in the training dataset. Then, the sample labels associated with the sample data are compared with the sample recognition results corresponding to the sample data with the same sequence. That is, according to the sample data sorting, the sample label associated with the first sample data is compared with the sample recognition result corresponding to the first sample data, and the loss value between the sample label and the sample recognition result is calculated. Then, the sample label associated with the second sample data is compared with the sample recognition result corresponding to the second sample data, and the loss value between the sample label and the sample recognition result is calculated. This process continues until the loss values ​​of all sample labels and all sample recognition results have been compared, and the predicted loss value can be obtained.

[0079] S704, when the predicted loss value reaches the preset convergence condition, the preset training model after convergence is determined as the image recognition model.

[0080] Understandably, the convergence condition can be either the predicted loss value being less than a set threshold, or the predicted loss value being very small after 50,000 calculations and no longer decreasing, at which point training can stop.

[0081] Specifically, when the predicted loss value fails to meet the preset convergence condition, the initial parameters of the preset training model are adjusted using the predicted loss value. The sample data and labels are then re-inputted into the adjusted preset training model, and the model is retrained to obtain a predicted loss value corresponding to that model. In this way, the prediction results continuously approach the correct result, and the accuracy of the preset training model increases until the predicted loss value reaches the preset convergence condition. At this point, the converged preset training model is determined as the image recognition model.

[0082] In this embodiment, by iteratively training the preset training model and calculating the overall loss value of the preset training model, the prediction loss value is determined, thereby determining the image recognition model and ensuring that the image recognition model has a high accuracy.

[0083] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0084] In one embodiment, a cardiac compression monitoring device is provided, which corresponds one-to-one with the cardiac compression monitoring method described in the above embodiments. For example... Figure 2 As shown, the cardiac compression monitoring device includes an initial image acquisition module 10, an initial compression operation module 20, an updated image acquisition module 30, and an updated compression operation module 40. Detailed descriptions of each functional module are as follows:

[0085] The initial image acquisition module 10 is used to monitor the target object corresponding to the establishment of extracorporeal membrane oxygenation (ECMO) in real time using an ultrasound device after the establishment of ECMO, and obtain the initial cardiac image corresponding to the target object.

[0086] The initial compression operation module 20 is used to acquire the image analysis result corresponding to the initial heart image, and input the image analysis result into the compression device, so that the compression device performs a compression operation on the target object according to the image analysis result;

[0087] The image acquisition module 30 is updated to acquire an updated cardiac image of the target object via an ultrasound device if the interval time reaches a preset interval period.

[0088] The update compression operation module 40 is used to obtain the update analysis result corresponding to the updated heart image, and input the update analysis result into the compression device, so that the compression device performs a compression operation on the target object according to the update analysis result.

[0089] In one embodiment, the initial press operation module 20 includes:

[0090] A model acquisition unit is used to acquire an image recognition model and input the initial heart image into the image recognition model;

[0091] An image recognition unit is used to perform image recognition on the initial heart image through the image recognition model to obtain at least one cross-sectional image;

[0092] The thrombus identification unit is used to perform thrombus identification on all the cross-sectional images respectively through the image recognition model, and obtain the image recognition result corresponding to each cross-sectional image;

[0093] The image parsing unit is used to determine the image parsing result corresponding to the initial heart image based on all the image recognition results.

[0094] In one embodiment, the thrombus identification unit includes:

[0095] A convolutional subunit is used to extract features from all the cross-sectional images through the convolutional layers in the image recognition model to obtain convolutional features corresponding to each cross-sectional image.

[0096] A pooling subunit is used to perform pooling processing on all the convolutional features through the pooling layer in the image recognition model to obtain pooling features corresponding to each of the cross-sectional images.

[0097] The recognition subunit is used to perform image recognition on all the pooled features through the fully connected layer in the image recognition model to obtain the image recognition result corresponding to each of the cross-sectional images.

[0098] In one embodiment, the initial press operation module 20 includes:

[0099] An information extraction unit is used to extract the probability of thrombosis from the image analysis result if the image analysis result indicates that the target object has a risk of thrombosis.

[0100] The parameter determination unit is used to determine the pressing depth, pressing frequency, and pressing time corresponding to the target object based on the formation probability.

[0101] The pressing operation unit is used to instruct the pressing device to perform a pressing operation on the target object according to the pressing depth, the pressing frequency and the pressing time.

[0102] In one embodiment, the initial image acquisition module 10 includes:

[0103] The image acquisition unit is used to control the ultrasound probe in the ultrasound device to detect at various preset positions in order to acquire images of different cardiac cross-sections of the target object and obtain an initial cardiac image containing multiple cross-sectional images.

[0104] In one embodiment, the device further includes:

[0105] The preset parameter acquisition unit is used to acquire a preset compression depth, a preset compression frequency, and a preset compression time when the image analysis result indicates that the target object has a risk of thrombosis, and input the preset compression depth, the preset compression frequency, and the preset compression time into the compression device;

[0106] The preset parameter execution unit is used to instruct the pressing device to perform a pressing operation on the target object according to the preset pressing depth, the preset pressing frequency, and the preset pressing time.

[0107] In one embodiment, the device further includes:

[0108] A sample acquisition module is used to acquire a training dataset, wherein the training dataset includes at least one sample data and a sample label corresponding to each of the sample data;

[0109] An image recognition module is used to acquire a preset training model, and to perform image recognition on each of the sample data through the preset training model to obtain sample recognition results.

[0110] The loss prediction module is used to determine the predicted loss value of the preset training model based on the sample label and the sample recognition result corresponding to the same sample data.

[0111] The model convergence module is used to determine the converged preset training model as an image recognition model when the predicted loss value reaches the preset convergence condition.

[0112] Specific limitations regarding the cardiac compression monitoring device can be found in the limitations of the cardiac compression monitoring method described above, and will not be repeated here. Each module in the aforementioned cardiac compression monitoring device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in hardware within or independently of the controller in the computer device, or stored in software in the memory of the computer device, so that the controller can invoke and execute the corresponding operations of each module.

[0113] In one embodiment, a computer device is provided, comprising a controller, a memory, a network interface, and a database connected via a system bus. The controller provides computing and control capabilities. The memory includes a readable storage medium and internal memory. The readable storage medium stores an operating system, computer programs, and a database. The internal memory provides an environment for the operation of the operating system and computer programs in the readable storage medium. The network interface is used to communicate with external terminals via a network connection. When executed by the controller, the computer program implements a cardiac compression monitoring method.

[0114] In one embodiment, a computer device is provided, including a memory, a controller, and a computer program stored in the memory and executable on the controller, wherein the controller executes the computer program to implement the cardiac compression monitoring method described above.

[0115] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a controller, implements the cardiac compression monitoring method described above.

[0116] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided by this invention can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0117] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is used as an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above.

[0118] The above-described embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.

Claims

1. A cardiac compression monitoring device, characterized by, include: The initial image acquisition module is used to monitor the target object corresponding to the establishment of extracorporeal membrane oxygenation (ECMO) in real time using ultrasound equipment after the establishment of ECMO, and obtain the initial cardiac image corresponding to the target object. The initial compression operation module is used to acquire the image analysis result corresponding to the initial heart image, and input the image analysis result into the compression device, so that the compression device performs a compression operation on the target object according to the image analysis result; An updated image acquisition module is used to acquire updated cardiac images of the target object via ultrasound equipment if the interval time reaches a preset interval period. The update compression operation module is used to obtain the update analysis result corresponding to the updated heart image, and input the update analysis result into the compression device, so that the compression device performs a compression operation on the target object according to the update analysis result; The initial press operation module includes: An information extraction unit is used to extract the probability of thrombosis from the image analysis result if the image analysis result indicates that the target object has a risk of thrombosis. The parameter determination unit is used to determine the pressing depth, pressing frequency, and pressing time corresponding to the target object based on the formation probability. The pressing operation unit is used to instruct the pressing device to perform a pressing operation on the target object according to the pressing depth, the pressing frequency and the pressing time.

2. A computer device comprising a memory, a controller and a computer program stored in the memory and executable on the controller, characterized in that, The controller is used to perform the following steps: After extracorporeal membrane oxygenation (ECMO) is established, the target object corresponding to the establishment of ECMO is monitored in real time using ultrasound equipment to obtain the initial cardiac image corresponding to the target object. Obtain the image analysis result corresponding to the initial heart image, and input the image analysis result into the compression device, so that the compression device performs a compression operation on the target object according to the image analysis result; If the interval reaches the preset interval period, an updated cardiac image of the target object is acquired by ultrasound equipment; Obtain the update analysis result corresponding to the updated heart image, and input the update analysis result into the compression device, so that the compression device performs a compression operation on the target object according to the update analysis result; The step of instructing the pressing device to perform a pressing operation on the target object based on the image analysis result includes: If the image analysis result indicates that the target object has a risk of thrombosis, then the probability of thrombosis is extracted from the image analysis result; Based on the formation probability, determine the pressing depth, pressing frequency, and pressing time corresponding to the target object; The pressing device is instructed to perform a pressing operation on the target object according to the pressing depth, the pressing frequency, and the pressing time.

3. The computer device as described in claim 2, characterized in that, The step of obtaining the image parsing result corresponding to the initial cardiac image includes: Obtain an image recognition model and input the initial heart image into the image recognition model; The initial heart image is subjected to image recognition using the image recognition model to obtain at least one cross-sectional image; The image recognition model is used to identify thrombi in all the cross-sectional images to obtain image recognition results corresponding to each cross-sectional image. Based on all the image recognition results, determine the image parsing result corresponding to the initial heart image.

4. The computer device as described in claim 3, characterized in that, The step of performing thrombosis identification on all the cross-sectional images using the image recognition model to obtain image recognition results corresponding to each cross-sectional image includes: The convolutional layer in the image recognition model is used to extract features from all the cross-sectional images to obtain convolutional features corresponding to each cross-sectional image. The pooling layer in the image recognition model is used to pool all the convolutional features to obtain pooled features corresponding to each of the cross-sectional images. The image recognition model uses a fully connected layer to perform image recognition on all the pooled features, thereby obtaining image recognition results corresponding to each of the cross-sectional images.

5. The computer device as described in claim 2, characterized in that, The step of using ultrasound equipment to monitor the target object corresponding to the establishment of extracorporeal membrane oxygenation in real time, and obtaining an initial cardiac image corresponding to the target object, includes: The ultrasound probe in the ultrasound device is controlled to probe at various preset positions to acquire images of different cardiac sections of the target object, thereby obtaining an initial cardiac image containing multiple cross-sectional images.

6. The computer device as described in claim 2, characterized in that, After obtaining the image parsing result corresponding to the initial heart image, the controller is further configured to perform the following steps: When the image analysis result indicates that the target object has a risk of thrombosis, a preset compression depth, a preset compression frequency, and a preset compression time are obtained, and the preset compression depth, the preset compression frequency, and the preset compression time are input into the compression device; The pressing device is instructed to perform a pressing operation on the target object according to the preset pressing depth, the preset pressing frequency, and the preset pressing time.

7. The computer device as described in claim 3, characterized in that, Before acquiring the image recognition model, the controller is also configured to perform the following steps: Obtain a training dataset, which includes at least one sample data and sample labels corresponding to each sample data; A preset training model is obtained, and image recognition is performed on each of the sample data using the preset training model to obtain the sample recognition result; The prediction loss value of the preset training model is determined based on the sample label and the sample recognition result corresponding to the same sample data. When the predicted loss value reaches the preset convergence condition, the preset training model after convergence is determined as the image recognition model.

8. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the controller, it performs the following steps: After extracorporeal membrane oxygenation (ECMO) is established, the target object corresponding to the establishment of ECMO is monitored in real time using ultrasound equipment to obtain the initial cardiac image corresponding to the target object. Obtain the image analysis result corresponding to the initial heart image, and input the image analysis result into the compression device, so that the compression device performs a compression operation on the target object according to the image analysis result; If the interval reaches the preset interval period, an updated cardiac image of the target object is acquired by ultrasound equipment; Obtain the update analysis result corresponding to the updated heart image, and input the update analysis result into the compression device, so that the compression device performs a compression operation on the target object according to the update analysis result; The step of instructing the pressing device to perform a pressing operation on the target object based on the image analysis result includes: If the image analysis result indicates that the target object has a risk of thrombosis, then the probability of thrombosis is extracted from the image analysis result; Based on the formation probability, determine the pressing depth, pressing frequency, and pressing time corresponding to the target object; The pressing device is instructed to perform a pressing operation on the target object according to the pressing depth, the pressing frequency, and the pressing time.

9. The computer-readable storage medium as claimed in claim 8, characterized in that, The step of obtaining the image parsing result corresponding to the initial cardiac image includes: Obtain an image recognition model and input the initial heart image into the image recognition model; The initial heart image is subjected to image recognition using the image recognition model to obtain at least one cross-sectional image; The image recognition model is used to identify thrombi in all the cross-sectional images to obtain image recognition results corresponding to each cross-sectional image. Based on all the image recognition results, determine the image parsing result corresponding to the initial heart image.

10. The computer-readable storage medium as claimed in claim 9, characterized in that, The step of performing thrombosis identification on all the cross-sectional images using the image recognition model to obtain image recognition results corresponding to each cross-sectional image includes: The convolutional layer in the image recognition model is used to extract features from all the cross-sectional images to obtain convolutional features corresponding to each cross-sectional image. The pooling layer in the image recognition model is used to pool all the convolutional features to obtain pooled features corresponding to each of the cross-sectional images. The image recognition model uses a fully connected layer to perform image recognition on all the pooled features, thereby obtaining image recognition results corresponding to each of the cross-sectional images.

11. The computer-readable storage medium as claimed in claim 8, characterized in that, The step of using ultrasound equipment to monitor the target object corresponding to the establishment of extracorporeal membrane oxygenation in real time, and obtaining an initial cardiac image corresponding to the target object, includes: The ultrasound probe in the ultrasound device is controlled to probe at various preset positions to acquire images of different cardiac sections of the target object, thereby obtaining an initial cardiac image containing multiple cross-sectional images.

12. The computer-readable storage medium as claimed in claim 8, characterized in that, After obtaining the image parsing result corresponding to the initial cardiac image, the computer program, when executed by the controller, further performs the following steps: When the image analysis result indicates that the target object has a risk of thrombosis, a preset compression depth, a preset compression frequency, and a preset compression time are obtained, and the preset compression depth, the preset compression frequency, and the preset compression time are input into the compression device; The pressing device is instructed to perform a pressing operation on the target object according to the preset pressing depth, the preset pressing frequency, and the preset pressing time.

13. The computer-readable storage medium as claimed in claim 9, characterized in that, Before acquiring the image recognition model, the computer program, when executed by the controller, also performs the following steps: Obtain a training dataset, which includes at least one sample data and sample labels corresponding to each sample data; A preset training model is obtained, and image recognition is performed on each of the sample data using the preset training model to obtain the sample recognition result; The prediction loss value of the preset training model is determined based on the sample label and the sample recognition result corresponding to the same sample data. When the predicted loss value reaches the preset convergence condition, the preset training model after convergence is determined as the image recognition model.

Images