Artificial intelligence data set construction method and apparatus related to cardiac contrast agent injection

Abstract

An artificial intelligence data set construction method and apparatus related to cardiac contrast agent injection. The method comprises: measuring an expiration pressure of a patient to be tested (600); measuring electrocardiogram changes of said patient; determining, by a control unit (500), an injection timing of a contrast agent injection unit (300) and sending a recording instruction for a hemodynamic condition in a heart; when a time point at which the expiration pressure reaches a first predetermined threshold, instantly monitoring, by the control unit, a sudden change amplitude of an electrocardiogram change rate; when the sudden change amplitude exceeds a change amplitude from a time point at which expiration starts to the time point at which the expiration pressure reaches the first predetermined threshold, determining a time point at which the sudden change amplitude occurs as a change inflection point of the electrocardiogram change rate; after determining the change inflection point of the electrocardiogram change rate, sending, by the control unit, an injection start signal to the contrast agent injection unit; and marking, by an artificial intelligence semantic training model (540) in the control unit, the electrocardiogram change rate of said patient sent by an electrocardiogram monitoring unit (200), thereby forming a training set of the electrocardiogram change rate.

Description

A method and apparatus for constructing an artificial intelligence dataset related to cardiac contrast agent injection. Technical Field

[0001] This invention relates to the field of intelligent medical technology, and in particular to a method and apparatus for constructing an artificial intelligence dataset related to cardiac contrast agent injection, and also to an auxiliary diagnostic device for patent foramen ovale based on the Valsalva maneuver. Background Technology

[0002] The foramen ovale is the physiological passageway of the atrial septum during embryonic development. After birth, due to increased left atrial pressure and decreased pulmonary artery resistance, the two parts of the atrial septum gradually fuse, and in most cases, it closes spontaneously within a year. If it fails to close, a patent foramen ovale (PFO) forms within the atrial septum. To diagnose PFO, provocation tests are typically used to increase diagnostic sensitivity. The Valsalva maneuver is the most commonly used provocation test, which significantly improves the diagnostic sensitivity of PFO by increasing intrathoracic pressure, reducing venous return, and increasing right atrial pressure. However, a full Valsalva maneuver is difficult to achieve in critically ill patients, intubated patients, and patients undergoing sedation-controlled transesophageal echocardiography (TEE).

[0003] In the diagnostic process, contrast echocardiography (ASCE) is an important tool. The criterion for diagnosing prolapsed fossa (PFO) via ASCE is the observation of contrast agent (ASC) passing through the foramen ovale on echocardiography. Specifically, if ASC is observed in the left atrium within 3–6 cardiac cycles after the right atrium is filled with it, this can also serve as a surrogate diagnostic indicator for PFO. This criterion is applicable when the shunt is not clearly visualized or is located in a non-current view, ensuring the accuracy and reliability of the diagnosis.

[0004] In the diagnosis of patent foramen ovale (PFO) using ASCE (contrast echocardiography), the accuracy of contrast agent injection directly impacts diagnostic accuracy, detection sensitivity, patient safety and comfort, as well as the utilization of medical resources and work efficiency. If the contrast agent fails to adequately reach the target area, even minute PFOs may go undetected. Non-standard injection can lead to uneven distribution of the contrast agent within the heart chambers, producing artifacts or false signals, resulting in misdiagnosis. Excessive contrast agent may increase the risk of adverse reactions, such as allergic reactions and kidney damage. To ensure high efficiency and reliability in diagnosis, strict control must be exercised over every step of the contrast agent injection process to ensure standardization and accuracy.

[0005] Accuracy in contrast agent injection involves five key aspects. First, the injection timing must be precise. The contrast agent must be injected during a specific phase of the heart (such as atrial systole) to ensure optimal diagnostic results. Second, the dosage must be accurate. The dosage of the contrast agent must be precisely calculated based on the patient's specific condition (such as weight and cardiac function) to avoid overdose or underdose. Third, the injection speed must be accurate. The injection speed must be moderate to ensure that the contrast agent is thoroughly mixed and evenly distributed. Fourth, the injection site must be accurate. The contrast agent should be injected into the appropriate vascular site (such as the right atrium) to avoid inaccurate diagnostic results due to injection site deviation. Fifth, real-time monitoring and feedback must be accurate. The distribution of the contrast agent needs to be monitored in real time during the injection process, and adjustments made as needed.

[0006] The accuracy of contrast agent injection has a significant impact on diagnostic results. First, accurate injection ensures uniform distribution of the contrast agent within the heart chambers, generating high-quality, clear ultrasound images. This helps physicians accurately observe and identify pinpoint foci (PFOs), reducing the possibility of misdiagnosis and missed diagnosis. Second, accurate contrast agent injection allows even minute PFOs to be more clearly visualized through the contrast agent's imaging effect, providing optimal contrast and improving the diagnostic performance of ultrasound imaging.

[0007] Furthermore, precise contrast agent injection can obtain high-quality diagnostic images in a single procedure, reducing the need for repeated examinations due to poor image quality, alleviating patient burden and discomfort. Simultaneously, by accurately controlling the dosage and injection rate of the contrast agent, the side effects and potential safety risks associated with overuse can be reduced. Precise injection avoids contrast agent waste, improves resource utilization efficiency, saves medical costs, and by obtaining accurate diagnostic results in a single procedure, avoids repetitive operations, improving the efficiency of the medical team and the smoothness of the diagnostic process.

[0008] Precise contrast agent injection ensures the reliability of diagnostic results, providing a solid basis for physicians to formulate appropriate treatment plans. Accurate diagnostic results can better guide subsequent interventional treatments or surgical procedures, improving treatment outcomes and safety. Accurate contrast agent injection helps obtain standardized imaging data, improving data consistency and comparability, which plays a crucial role in subsequent patient management and research analysis. A precise injection process reduces unnecessary repeated examinations and procedures, minimizing patient pain and anxiety, and increasing patient satisfaction.

[0009] US2016213833(A1) discloses a method for determining the optimal volume of contrast agent to be administered to a patient, comprising the steps of: (a) measuring one or more patient-specific physiological parameters, including at least the patient's heart rate; and (b) determining the optimal volume of contrast agent based on the patient-specific physiological parameters determined in (a). This technical solution focuses on determining the optimal contrast agent volume based on patient-specific physiological parameters (such as heart rate). While volume determination is important, the need for injection accuracy in ASCEs extends beyond volume, encompassing injection time, speed, and location. Simply determining the volume is insufficient to ensure uniform distribution of the contrast agent in the heart and optimal imaging. ASCEs require real-time monitoring of the contrast agent distribution during injection and adjustments as needed to ensure high-quality images and accurate diagnosis. However, the technical solution does not mention a mechanism for real-time monitoring and feedback, which is insufficient for the delicate procedure of ASCE diagnosis of patent foramen ovale (PFO).

[0010] US2014206991(A1) discloses a method for predicting anticipated contrast agent changes (CEP) for contrast agent-assisted examinations, as shown in Figure 10. The method includes the following steps: acquiring patient-specific blood flow signals in the predicted quantity under defined examination protocols, and determining a separate impulse response function from these signals; acquiring heart rate signals corresponding to the predicted quantity and obtaining heart rate data; and predicting contrast agent changes based on the separate impulse response function, average cardiac rate (CA), the current examination protocol, and the currently acquired heart rate data. This technical solution predicts contrast agent behavior based on patient-specific blood flow signals, heart rate signals, and the current examination protocol. The CEP algorithm includes extracting the patient's function from the blood flow signal using the AIR algorithm, obtaining a corrected patient function, and adjusting the corrected patient function to the PRED algorithm based on the current examination protocol. The PRED algorithm is then used to finally obtain the predicted contrast agent changes. However, ASCE requires real-time monitoring and adjustment during actual operation to ensure accurate injection of the contrast agent into the target area and avoid artifacts and erroneous signals. Relying solely on prediction may not be sufficient to handle the complexities and variations encountered in actual practice. ASCE diagnostic procedures also require precise control of injection time, speed, dosage, and location to ensure the contrast agent fully penetrates the target area and distributes evenly. While this technical solution provides a predictive mechanism, it does not specifically address how to precisely control these parameters in actual operation.

[0011] Therefore, the aforementioned existing technologies are not applicable to current ASCE. Therefore, for ASCE, this invention aims to provide a method and apparatus for constructing an artificial intelligence dataset related to cardiac contrast agent injection, hoping to significantly improve the effectiveness and efficiency of ASCE in diagnosing patent foramen ovale (PFO), thereby optimizing injection parameters, providing real-time monitoring and feedback, improving diagnostic accuracy and sensitivity, and reducing patient pain and discomfort. This dataset also provides physicians with a powerful tool to help them make more accurate diagnoses and develop more effective treatment plans.

[0012] Summary of the Invention

[0013] Currently, transesophageal echocardiography combined with right ventricular contrast echocardiography and adequate provocation testing is the gold standard imaging method for diagnosing patent foramen ovale (PFO). However, there is no unified standard for this procedure and diagnostic criteria in China. Medical staff often instruct patients to exhale and hold their breath based on experience, determining the duration of each action. This leads to inconsistencies in the duration of the Valsalva maneuver, and the timing of intravenous contrast agent injection and echocardiogram imaging is also determined by experience. Consequently, some echocardiogram images fail to accurately reflect cardiac hemodynamics, resulting in misdiagnosis of PFO.

[0014] Existing technologies have developed solutions that use a Valsalva respiration detection mechanism to monitor whether a patient's Valsalva respiration is within acceptable limits and a contrast agent mixing and injection mechanism to mix the contrast agent and maintain its activity. For example, CN114272470A discloses a foramen ovale screening foaming injector, comprising: a Valsalva respiration detection mechanism for detecting whether a patient's Valsalva respiration is within acceptable limits; a receiving module for receiving external injection commands; a contrast agent mixing and injection mechanism for mixing and injecting the contrast agent; and a control module for controlling the contrast agent mixing and injection mechanism based on the Valsalva respiration detection mechanism and the receiving module. The control module controls the contrast agent mixing and injection mechanism to mix the contrast agent. After the Valsalva respiration detection mechanism detects that the patient's Valsalva respiration is within acceptable limits and the receiving module receives an external injection command, the control module controls the contrast agent mixing and injection mechanism to perform the injection. When the expiratory pressure sensor in the Valsalva respiration detection mechanism detects that the expiratory pressure reaches a preset pressure, it activates a corresponding timing device to record the duration for which the expiratory pressure is maintained. When this duration reaches a preset time, it is determined that the patient's Valsalva respiration is within acceptable limits. However, the judgment of patient test parameters in this technical solution is limited to the continuous process of the state. According to this threshold setting method, the test parameters that are deemed qualified are the results after the continuous changes of the patient's physical indicators for a period of time. They cannot reflect the continuous changes of the patient's physical state (such as cardiac hemodynamics). If the timing of contrast agent injection or cardiac ultrasound imaging is selected based solely on the data at the end of the physiological change characteristics, the contrast agent will not be able to match the most suitable cardiac hemodynamics or the timing of cardiac ultrasound imaging will lag behind the actual shunt in the heart, thus making the test results inaccurate. In other words, this technical solution only judges the standard of a patient's performance of the Valsalva maneuver based on the form of the maneuver, i.e., whether the patient has performed the maneuver according to the current assessment standards. However, it cannot judge whether the same standard Valsalva maneuver will produce the same physiological changes in different patients. For example, due to individual differences among patients, the same Valsalva maneuver will usually produce different physiological changes in different patients. The patient's physiological changes (such as cardiac hemodynamics) are directly related to the accurate selection of the timing of contrast agent injection. The appropriate injection timing can ensure that the contrast agent reaches the highest concentration in a specific part of the heart, thereby ensuring that the contrast agent can achieve more accurate imaging feedback under the corresponding physiological changes, thus obtaining clear images and helping doctors make accurate diagnoses.

[0015] This invention aims to provide an auxiliary diagnostic device for patent foramen ovale based on the Valsalva maneuver to accurately determine and alert the timing of cardiac shunt and contrast agent administration, thereby generating accurate ultrasound images of intracardiac shunts and providing accurate diagnostic evidence for patent foramen ovale.

[0016] To address the shortcomings of existing technologies, this invention provides an auxiliary diagnostic device for patent foramen ovale based on the Valsalva maneuver, comprising a blowing unit, an electrocardiogram (ECG) monitoring unit, a contrast agent injection unit, and a control unit. The blowing unit measures the blowing pressure of the patient; the ECG monitoring unit measures the ECG changes of the patient; the contrast agent injection unit administers contrast agent to the patient; when the blowing pressure of the patient, as reported by the blowing unit, reaches a first predetermined threshold and remains at that threshold for a predetermined duration, the control unit begins to request the ECG rate of change of the patient output by the ECG monitoring unit. This invention utilizes the control unit and a pressure sensor to collect data on the blowing pressure and duration, thereby providing an accurate timing for judging ECG changes and avoiding repeated Valsalva maneuvers for the patient.

[0017] When the rate of change of the patient's electrocardiogram (ECG) is determined to reach an inflection point related to the Valsalva maneuver, the control unit sends a prompt to the contrast agent injection unit to administer contrast agent. This invention analyzes the ECG rate of change and determines the timing of the inflection point, enabling the contrast agent injection unit to promptly inject contrast agent intravenously, avoiding missed injection opportunities, and standardizing the timing of contrast agent injection.

[0018] In response to the received start information of contrast agent intravenous infusion from the contrast agent injection unit, the control unit issues a breathing command to the patient to be tested through the prompting unit, and when it is determined that the breathing pressure of the patient to be tested measured by the breathing unit is lower than a second predetermined threshold, the control unit issues a recording command of the hemodynamic status in the heart through the prompting unit.

[0019] Unlike existing technologies, the auxiliary diagnostic device of this invention can measure the electrocardiogram (ECG) changes of the patient under test through an ECG monitoring unit to monitor the ECG change rate when the patient's breathing pressure reaches a first predetermined threshold and remains at that threshold for a predetermined duration. When the ECG change rate of the patient under test reaches an inflection point related to the Valsalva maneuver, the control unit sends a prompt to the contrast agent injection unit to administer contrast agent. Simultaneously, the control unit can issue a recording command for intracardiac hemodynamics based on the initiation information of intravenous contrast agent administration via a prompting unit. Based on these distinguishing technical features, the problems to be solved by this invention include: how to improve the accuracy of determining the timing of contrast agent injection and how to improve the accuracy of determining the timing of cardiac ultrasound imaging, so as to generate and obtain accurate ultrasound images of intracardiac shunts, providing accurate diagnostic evidence for patent foramen ovale. The contrast agent injection unit of this invention is connected to the control unit. The control unit prompts for contrast agent injection based on the monitoring results of ECG change rate and breathing pressure, thereby enabling precise control of the contrast agent injection timing based on real-time data of ECG changes and breathing pressure, improving diagnostic accuracy. To address the shortcomings of relying on work experience to determine the timing of ultrasound imaging and the inability to directly observe the patient's breathing patterns, this invention can determine the timing of ultrasound imaging based on changes in the breathing patterns, thereby capturing the shunt within the heart in a timely manner and avoiding missing the optimal imaging opportunity.

[0020] Compared to existing technologies that only measure expiratory pressure while ignoring the correlation analysis between the patient's actual Valsalva maneuver and the rate of change of electrocardiogram (ECG), this invention can determine the intracardiac hemodynamic conditions suitable for contrast agent injection by analyzing the temporal correlation between the ECG rate of change and the Valsalva maneuver. In this case, the intracardiac hemodynamic conditions suitable for contrast agent injection are characterized by the inflection point of the ECG rate of change in relation to the Valsalva maneuver over time. Preferably, the determination of the inflection point of the ECG rate of change, particularly in relation to the Valsalva maneuver over time, is performed as follows:

[0021] S01: Initial monitoring of ECG rate of change. Monitoring of the ECG rate of change begins at least from the start of breathing. The ECG monitoring unit is activated immediately upon the patient initiating the Valsalva maneuver (breathing action) to begin real-time monitoring of the ECG rate of change. The control unit receives the ECG signal from the monitoring unit and performs preliminary analysis using an artificial intelligence semantic training model.

[0022] S02: Monitoring of ECG rate of change mutations after the ventilation pressure reaches the target. Monitoring of the mutation amplitude of the ECG rate of change begins at the time point when the ventilation pressure reaches the first predetermined threshold. The ventilation unit detects that the ventilation pressure has reached the first predetermined threshold (e.g., 40 mmHg), and this information is immediately transmitted to the control unit. The control unit then switches to a more intensive monitoring mode, focusing on the mutation amplitude of the ECG rate of change.

[0023] S03: Identify the inflection point of the electrocardiogram (ECG) rate of change. When the abrupt change exceeds the range of change from the start of exhalation to the point when the exhalation pressure reaches the first predetermined threshold, the point in time where the abrupt change occurs is determined as the inflection point. The control unit analyzes the ECG rate of change data from the start of exhalation to the point when the pressure reaches the target, searching for the abrupt change. Once the detected abrupt change exceeds the maximum range of change during the previously monitored period, and the extent of the exceedance reaches or exceeds a preset percentage (e.g., more than 30%, preferably more than 50%, and especially preferably more than 70%), the point in time where the abrupt change occurs is determined as the inflection point of the ECG rate of change related to the Valsalva maneuver.

[0024] S04: Preparation for contrast agent injection. After determining the inflection point of the electrocardiogram rate of change, the control unit immediately sends a preparation message to the contrast agent injection unit. Medical staff, following the instructions from the control unit, prepare to inject the contrast agent into the patient's vein.

[0025] S05: Synchronizing Contrast Agent Injection with ECG Changes. After confirming that the ECG rate of change meets preset conditions, the control unit sends an injection start signal to the contrast agent injection unit. While medical staff perform the injection, the control unit continues to monitor the ECG rate of change to ensure that the injection timing is synchronized with the inflection point of the ECG rate of change.

[0026] S06: Recording cardiac blood flow dynamics. After injection, the control unit issues an exhalation command to the patient via the prompting unit and instructs the ultrasound imaging unit to begin recording the dynamic blood flow within the heart. The ultrasound imaging unit captures and records changes in cardiac blood flow, providing crucial information for diagnosis.

[0027] According to a preferred embodiment, when a blowing command is issued, the control unit generates labeled prompt information through a prompting unit. The artificial intelligence semantic training model within the control unit labels the rate of change of the patient's electrocardiogram (ECG) sent by the ECG monitoring unit, thereby forming a training set of ECG rate of change. By collecting relevant data on ECG changes and forming a training set, the artificial intelligence semantic training model can learn and identify inflection points of change related to the Valsalva maneuver based on the training set.

[0028] According to a preferred embodiment, the control unit integrates an artificial intelligence semantic training model. After training, this model analyzes the rate of change in the patient's electrocardiogram (ECG) data sent by the ECG monitoring unit and determines the inflection point related to the Valsalva maneuver. This invention achieves accurate reminders of contrast agent infusion time by training an artificial intelligence semantic training model to analyze ECG rate of change and promptly determine inflection points in heart rate.

[0029] According to a preferred embodiment, when the control unit issues a blowing command via a prompting unit, the blowing unit monitors the blowing pressure and its changes in the tubing. Based on the information condition that the blowing pressure reaches a first predetermined threshold, the control unit triggers a timing command to start the connected timer. When the blowing pressure reaches the first predetermined threshold for a predetermined duration, the control unit generates a command to instruct the patient to hold their breath and a timing prompt for the duration of breath-holding. The control unit of this invention can monitor changes in blowing pressure and duration through a timer and pressure sensor, and can also prompt the patient to cooperate in achieving standardized blowing pressure, thus standardizing the patient's Valsalva maneuver. This reduces the likelihood of non-standard Valsalva maneuvers and improves the efficiency of the diagnosis of patent foramen ovale.

[0030] According to a preferred embodiment, in response to the end information of the contrast agent intravenous infusion from the contrast agent injection unit, the control unit issues an exhalation command to the patient through a prompting unit and determines whether the blowing pressure of the patient measured by the blowing unit is lower than a second predetermined threshold. Unlike the prior art, the control unit of the present invention can issue corresponding breathing commands to the patient through the prompting unit based on the contrast agent intravenous infusion process. Based on the above-mentioned distinguishing technical features, the problem to be solved by the present invention may include: how to reduce the difficulty for the patient to perform the Valsalva maneuver without affecting the accuracy of the test results. Specifically, without clear instructions, the patient may hold their breath or inhale prematurely due to uncertainty about whether they can breathe, which can easily lead to dizziness, hypoxia, and non-standard Valsalva maneuvers after the procedure. The present invention issues clear instructions by monitoring the blowing pressure, making it easier for the patient to cooperate and reducing the difficulty of performing the Valsalva maneuver correctly.

[0031] According to a preferred embodiment, the blowing unit includes a pressure sensor, a pressure communication port, a blowing assembly, and a conduit. The pressure sensor is disposed within the conduit connected to the blowing assembly to collect the blowing pressure. The pressure sensor transmits the blowing pressure data to the control unit via the pressure communication port. This invention monitors the blowing pressure and its changes using a control unit and a pressure sensor, which facilitates timely issuance of instructions for the correct timing of blowing, inhalation, and breath-holding actions, standardizing the Valsalva maneuver and improving the effectiveness of echocardiographic imaging.

[0032] According to a preferred embodiment, the electrocardiogram (ECG) monitoring unit measures changes in heart rate, P waves, and QRS complexes. The ECG monitoring unit compares the ECGs of the patient before and after performing the Valsalva maneuver to measure these changes. By comparing ECGs before and after the Valsalva maneuver, this invention allows for more timely assessment of heart rate changes and the identification of inflection points in the heart rate.

[0033] According to a preferred embodiment, the device further includes a display unit. Based on changes in the blowing pressure, the control unit controls the display unit to display the changes in blowing pressure across different pressure ranges using indicator changes, thereby alerting the patient or medical staff to the current blowing status. This helps the patient and medical staff to understand the current blowing status based on the indicator changes, thus encouraging them to actively cooperate in completing the diagnostic process.

[0034] According to a preferred embodiment, the control unit synchronously displays the blowing pressure and its changes, timing information, and electrocardiogram changes via a display unit. This invention integrates these three types of information into a single display unit, enabling medical personnel to simultaneously monitor changes in all types of information and promptly perform various operations based on received prompts, thus standardizing the diagnostic procedures for patent foramen ovale.

[0035] According to a preferred embodiment, when the patient's heart rate slows down and reaches a preset heart rate threshold, the control unit sends a prompt to the contrast agent injection unit to administer contrast agent. Reaching the preset heart rate threshold indicates an inflection point in heart rate change. Accurate judgment and alerting at this point prevent medical staff from missing the optimal time for ultrasound imaging, thus enabling the acquisition of optimal cardiac blood flow changes and providing the most accurate diagnostic information.

[0036] According to a preferred embodiment, during the preparation of the contrast agent, the contrast agent is uniformly mixed with air by repeatedly pulling the syringe to form a homogeneous mixture of air and contrast agent. The advantages of this approach are: it improves the quality of the captured images, making the contours of internal structures clearer; it reduces the need for contrast agent, lowering the risk of side effects and allergic reactions experienced by the patient; the air-mixed contrast agent is more evenly distributed and may also have lower viscosity, making the mixture easier to inject and deliver.

[0037] This invention provides, from a third aspect, a processor for constructing an artificial intelligence dataset related to cardiac contrast agent injection. The processor incorporates an artificial intelligence semantic training model. First, the processor issues a breath-blowing command to the patient via a connected prompting unit (potentially a speech synthesizer or display screen). This command may be a voice prompt informing the patient of the need for a breath-blowing test, or a text prompt displayed on a screen for the patient to read. When the patient blows as instructed, the breath-blowing unit connected to the processor measures and transmits the pressure data generated by the patient's breath. Simultaneously, an electrocardiogram (ECG) monitoring unit monitors and transmits the patient's ECG changes. In this step, the prompting unit generates labeled prompt information based on the instructions sent by the processor. This includes the breath pressure reading, a graphical or numerical representation of the ECG changes, and warnings for any abnormalities. The artificial intelligence semantic training model analyzes the received ECG change data. Based on the labeled information of the ECG rate of change, the artificial intelligence semantic training model learns to determine the inflection point of the ECG rate of change. The labeled information of the ECG rate of change contains a large amount of ECG change data and corresponding labels, used to train the artificial intelligence semantic training model to recognize and predict heart rate changes. By analyzing the rate of change in electrocardiogram (ECG), the AI ​​semantic training model can promptly identify inflection points in heart rate changes. When the AI ​​semantic training model identifies an inflection point in heart rate changes, it can alert medical staff to administer contrast agents at the optimal time.

[0038] By connecting the processor to the ventilation unit and ECG monitoring unit, real-time data collection and annotation are achieved, ensuring high accuracy and completeness of the data, thereby improving the quality of the dataset. Combining ventilation pressure and ECG changes, and integrating data from multiple dimensions, helps to build a more comprehensive and complex dataset, providing a foundation for training more powerful AI models.

[0039] According to a preferred embodiment, during the exhalation process, when the exhalation pressure reaches a first predetermined threshold, the processor instantly monitors the abrupt change in the electrocardiogram (ECG) rate of change. To determine the abrupt change amplitude, the processor first begins collecting ECG and exhalation pressure data from the start of exhalation, typically at a frequency of 100 times per second, to ensure data accuracy and precision. Next, the processor calculates the initial amplitude of the ECG rate of change during the period from the start of exhalation to the exhalation pressure reaching the first predetermined threshold. This process uses statistical methods, such as mean and standard deviation, to calculate the ECG rate of change during this period, which is defined as the initial amplitude of change.

[0040] After the blowing pressure reaches a first predetermined threshold, the processor monitors the changes in the ECG rate of change in real time and compares the current change in the ECG rate of change with the initial change. If the sudden change in the current ECG rate of change exceeds a predetermined proportion of the initial change (e.g., 1.5 times or 2 times), the processor confirms a significant mutation. At this point, the processor uses a mutation detection algorithm (such as CUSUM or Bayesian mutation detection) to reduce the impact of short-term fluctuations and ensure the accuracy of the judgment. Once a significant mutation is identified, the processor records the precise time point of the event and marks it as the inflection point of the ECG rate of change, storing it in memory for subsequent analysis and verification.

[0041] During ventilation, the variation in ventilation pressure is also a key monitoring indicator. When ventilation begins, the patient follows instructions to exhale, and the ventilation unit monitors and records the changes in ventilation pressure in real time using high-precision sensors. Throughout the process, the processor acquires ventilation pressure data at a high frequency (e.g., 100 times per second) and transmits this data to the processor for analysis via wireless or wired connection. The processor receives and analyzes the ventilation pressure data in real time, while using smoothing algorithms (such as moving average or Kalman filtering) to reduce data noise and ensure the accuracy of the analysis results.

[0042] During the period from the start of ventilation to the point where the ventilation pressure reaches a first predetermined threshold, the processor calculates and records the initial change in ventilation pressure. When the ventilation pressure reaches the predetermined threshold, the processor records this time point and switches to ECG variability monitoring mode to continue monitoring ventilation pressure and ECG data at a high frequency. Afterward, the processor continuously monitors and records changes in ventilation pressure for subsequent analysis and verification.

[0043] When the processor determines the inflection point of the rate of change in ECG variability, i.e., the time point of the abrupt change in amplitude, it generates a labeled alert. This information includes detailed information such as the time point of the abrupt change, ECG variability data, the amplitude of the abrupt change, and ventilation pressure data. The processor transmits this information to the alert unit, which promptly conveys it to medical staff through screen display, voice announcement, or vibration alerts. Simultaneously, the processor records and stores all generated labeled alerts in internal or external memory, ensuring data integrity and traceability, and regularly backs them up to cloud storage or external hard drives to prevent data loss. This series of detailed steps ensures the processor's operational accuracy and efficiency in monitoring ventilation pressure and ECG variability, as well as the timeliness and reliability of the labeled alerts, providing strong support for the medical process.

[0044] According to a preferred embodiment, during the blowing process, when the blowing pressure reaches a first predetermined threshold, the processor triggers a corresponding timer to start timing. First, the processor monitors the blowing pressure data collected by the blowing unit in real time, typically at a frequency set multiple times per second (e.g., 100 times per second) to ensure data continuity and accuracy. Then, a first predetermined threshold (e.g., 10 cmH2O) is preset and displayed and confirmed on the processor's control panel or software interface. When the processor detects that the real-time blowing pressure reaches or exceeds this predetermined threshold, it generates a timing command, causing the connected timer to start timing and recording the precise time point from when the blowing pressure reaches the first predetermined threshold.

[0045] In the next steps, the processor generates instructions to guide the patient in holding their breath. First, a predetermined duration (e.g., 30 seconds) is set according to clinical requirements or experimental design, and this duration is displayed and confirmed on the processor's control panel or software interface. The processor monitors the timing progress in real time after the ventilation pressure reaches a first predetermined threshold, comparing the timer's reading with the predetermined duration. When the timer reaches the predetermined duration, the processor generates instructions to ask the patient to begin holding their breath, specifying the exact duration. Additionally, the processor generates a prompt indicating the breath-holding duration, including the specific time (e.g., 10 seconds), and communicates this information to the patient or healthcare personnel through a prompt unit or similar means.

[0046] To ensure precise timing control of data collection, the processor employs a high-precision timer to minimize timing errors and records the time points of all key events in real time, including when the ventilation pressure reaches a predetermined threshold, the timing begins, and the predetermined duration is reached. All collected data, including time points and corresponding ventilation pressures, ECG variability rates, etc., are recorded and stored in internal or external memory for subsequent time series analysis. By setting the timer and predetermined duration, precise timing control of data collection is ensured, providing a reliable foundation for centralized time series analysis of the data. The processor flexibly adjusts the data collection process based on dynamic changes in ventilation pressure and predetermined duration, improving data diversity and coverage.

[0047] According to a preferred embodiment, the processor and memory are connected via wired or wireless means. For example, the processor connects to the memory through a high-speed bus interface (such as SPI, I2C, or PCIe) to ensure data transmission speed and stability. The memory is used to store various types of data information acquired and generated by the processor. This data information may include various types of received data, such as the rate of change of electrocardiogram (ECG) and its inflection point, breath-holding duration, time and pressure related to ventilation, and injection time of the contrast agent injection unit, and may also include various types of instruction data. The processor formats the acquired data information into a structured data format suitable for storage (such as JSON, XML, or CSV). The processor generates a timestamp for each record and establishes a time index to facilitate subsequent chronological data retrieval. The processor marks events such as the inflection point of the rate of change of ECG, breath-holding duration, key time points of ventilation pressure, and contrast agent injection time, and establishes an event index. The processor integrates the contrast agent injection time information into the overall data record and stores it in association with information such as the rate of change of ECG and ventilation pressure. The processor marks the injection time information to facilitate subsequent data analysis and retrieval, ensuring data integrity and traceability, and providing sufficient raw data for subsequent data analysis and model training. Attached Figure Description

[0048] Figure 1 is a schematic diagram of the hardware connection relationship of the auxiliary diagnostic device for patent foramen ovale provided by the present invention;

[0049] Figure 2 is a schematic diagram of the logical connection relationship of the auxiliary diagnostic device for patent foramen ovale provided by the present invention;

[0050] Figure 3 is a schematic diagram of the display unit provided by the present invention;

[0051] Figure 4 is a schematic diagram of the steps of the contrast agent application method provided by the present invention;

[0052] Figure 5 is a logical diagram of the diagnostic method for patent foramen ovale based on the Valsalva maneuver;

[0053] Figure 6 is a schematic diagram of one of the PFO diagnostic images of patient A;

[0054] Figure 7 is a schematic diagram of another PFO diagnostic image of patient A;

[0055] Figure 8 is a schematic diagram of one of the PFO diagnostic images of patient B;

[0056] Figure 9 is a schematic diagram of the processor connection for constructing an artificial intelligence dataset related to cardiac contrast agent injection provided by the present invention.

[0057] Figure 10 is a schematic diagram of another auxiliary diagnostic device for patent foramen ovale in the prior art.

[0058] List of reference numerals: 100: Inhalation unit; 110: Pressure sensor; 120: Pressure communication port; 130: Inhalation assembly; 140: Tube; 200: ECG monitoring unit; 210: ECG communication port; 220: Electrode; 230: Sensing module; 300: Contrast agent injection unit; 310: Contrast agent; 320: Syringe; 330: Actuation component; 340: Injection chamber; 350: Injection communication port; 400: Ultrasound imaging unit; 410: Ultrasound probe; 420: Imaging communication port; 500: Control unit; 510: Prompt unit; 520: Control communication port; 530: Processor; 540: Artificial intelligence semantic training model; 550: Memory; 560: Timer; 600: Patient to be tested; 700: Display unit. Detailed Implementation

[0059] The following is a detailed explanation with reference to the accompanying drawings.

[0060] The Valsalva maneuver is a clinical physiological test in which the patient performs a forceful exhalation-closing action, that is, after taking a deep breath, the patient closes the glottis tightly and then forcefully exhales, resisting the closed epiglottis during exhalation. This increases intrathoracic pressure and affects blood circulation and autonomic nervous system function, thereby achieving diagnostic and therapeutic purposes.

[0061] In this invention, a mixture of physiological saline and air is selected as contrast agent 310.

[0062] A mixture of physiological saline and air can be used as a contrast agent 310 for echocardiography, particularly in assessing right ventricular function and detecting patent foramen ovale (PFO). This contrast agent 310 is commonly referred to as "saline contrast" or "acoustic contrast," and it enhances ultrasound images by rapidly injecting physiological saline containing microbubbles.

[0063] In this invention, as shown in Figure 4, the method for applying "physiological saline contrast imaging" or "acoustic contrast imaging" includes the following steps:

[0064] S11: Preparation steps: Take a certain amount of physiological saline, for example, 9 ml of physiological saline. The specific volume of physiological saline is as needed and is not limited here.

[0065] S12: Mixing Step: Mix 9 ml of physiological saline with 1 ml of air to form a solution containing microbubbles. This can be achieved by rapidly and repeatedly pushing and pulling the two 10 ml spiral syringes connected by the three-way connector 20-30 times to ensure thorough mixing of the physiological saline and air.

[0066] Preferably, during the preparation of the contrast agent 310, the contrast agent 310 is uniformly mixed with air by repeatedly pulling the syringe 320 to form a uniform mixture of air and contrast agent 310.

[0067] S13: Intravenous injection procedure: Quickly inject the microbubble mixture into a vein, usually the antecubital vein.

[0068] S14: Monitoring procedure: Immediately after injection, use an ultrasound device to observe and assess the angiography results and cardiac structure.

[0069] Example 1

[0070] Currently, transesophageal echocardiography (TEE) combined with right ventricular contrast echocardiography (ASCE) and a thorough provocation test is the gold standard imaging method for diagnosing patent foramen ovale (PFO). However, there is no unified standard for this procedure and diagnostic criteria in China. Medical staff, based on their experience, instruct patients to exhale and hold their breath, determining the duration of each action. This results in inconsistent durations of the Valsalva maneuver performed by patients. Furthermore, the timing of intravenous injection of contrast agent 310 and the timing of echocardiogram imaging are all determined by experience. Consequently, some echocardiogram images fail to accurately reflect cardiac hemodynamics, leading to misdiagnosis of PFO.

[0071] The present invention aims to accurately determine and alert the timing of cardiac shunt and contrast agent 310 administration by providing a Valsalva maneuver-based auxiliary diagnostic device for patent foramen ovale, thereby generating accurate ultrasound images of intracardiac shunts and providing accurate diagnostic evidence for patent foramen ovale.

[0072] To address the shortcomings of existing technologies, this invention provides an auxiliary diagnostic device for patent foramen ovale based on the Valsalva maneuver, as shown in Figure 1, comprising an air blowing unit 100, an electrocardiogram monitoring unit 200, and a contrast agent injection unit 300. As shown in Figure 2, the air blowing unit 100, the electrocardiogram monitoring unit 200, and the contrast agent injection unit 300 are communicatively connected to the control unit 500.

[0073] As shown in Figures 1 and 2, the blowing unit 100 is used to measure the blowing pressure of the patient 600. Preferably, the blowing unit 100 includes a pressure sensor 110, a pressure communication port 120, a blowing assembly 130, and a conduit 140. The pressure sensor 110 is disposed in the conduit 140 connected to the blowing assembly 130 to collect the blowing pressure. The pressure sensor 110 sends the blowing pressure data to the control unit 500 through the pressure communication port 120. This invention monitors the blowing pressure and its changes through the control unit 500 and the pressure sensor 110, which is more conducive to issuing timely instructions on the correct timing of blowing, inhalation, and breath-holding actions, achieving standardization of the Valsalva maneuver, and improving the effectiveness of cardiac ultrasound imaging.

[0074] As shown in Figures 1 and 2, the ECG monitoring unit 200 is used to measure the ECG changes of a patient 600. The ECG monitoring unit 200 includes an ECG communication port 210, a sensing module 230, leads, and electrodes 220. The sensing module 230 is located within the monitoring and display component box and is connected to the electrodes 220 via leads. The sensing module 230 is primarily responsible for sensing the ECG signals of the patient 600. It typically consists of one or more electrodes 220 connected to the patient's body via leads, which contact the patient's body to detect changes in the ECG signals. The electrodes 220 can be attached to the radial artery of the patient 600 using adhesive tape. The electromyography (EMG) signals collected by the electrodes 220 are used to construct an electrocardiogram (ECG), and the ECG waveform is displayed on the display unit 700. The sensing module 230 communicates with the control unit 500 via the ECG communication port 210.

[0075] For example, when the heart contracts, the generated electromyographic signals are transmitted to the sensing module 230 via electrodes 220 and wires. The sensing module 230 receives and amplifies these electrical signals and converts them into digital signals. The digital signals are then transmitted to the ECG communication port 210, and through the connected control unit 500, to the display unit 700 for monitoring and observation by medical personnel.

[0076] As shown in Figures 1 and 2, the contrast agent injection unit 300 is used to administer contrast agent 310 to the patient 600. The contrast agent injection unit 300 includes the contrast agent 310 and a syringe 320. Preferably, the syringe 320 consists of two 10 ml spiral syringes connected in a three-way configuration, allowing medical personnel to manually inject the contrast agent 310. When the syringe 320 can be controlled by the control unit 500, it also includes an actuator 330 capable of performing push-pull movements. The contrast agent 310 is filled within the injection cavity 340 of the syringe 320. The actuator 330 is communicatively connected to the control unit 500 via an injection communication port 350. The actuator 330 is, for example, a displacement sensor mounted on the plunger of the syringe 320. The displacement sensor transmits displacement data to the control unit 500 via the injection communication port 350, enabling the control unit 500 to determine the injection action and injection stop time of the syringe 320 based on the displacement data.

[0077] As shown in Figures 1 and 2, the ultrasound imaging unit 400 is used to image the hemodynamics within the heart of the patient 600 to monitor right-to-left intracardiac shunt at the atrial level. The ultrasound imaging unit 400 includes an ultrasound imaging subject, an ultrasound probe 410, and an image communication port 420. The ultrasound imaging subject transmits information to the control unit 500 via the image communication port 420, indicating whether the imaging action has been completed.

[0078] As shown in Figure 2, when the blowing pressure of the patient 600, as fed back by the blowing unit 100, reaches a first predetermined threshold and remains there for a predetermined duration, the control unit 500 begins to request the electrocardiogram (ECG) rate of change of the patient 600 output by the ECG monitoring unit 200. This invention utilizes the control unit 500 and the pressure sensor 110 to collect data on the blowing pressure and duration, thereby providing an accurate timing for judging ECG changes and preventing the patient 600 from repeatedly performing the Valsalva maneuver.

[0079] For example, to ensure the accuracy and effectiveness of the examination process, the patient 600 needs to receive detailed instructions before undergoing cardiac angiography to correctly perform the Valsalva maneuver. Specifically, the patient 600 needs to perform this specific procedure under the guidance of a medical professional: first, take a deep breath, then forcefully attempt to exhale, while keeping the airway closed, even if the gas cannot actually be expelled. This forceful exhalation, similar to straining to defecate or holding one's breath, creates pressure and flow of the gas, generating pressure within tubing 140. When the gas pressure within tubing 140 reaches 40 mmHg, the duration of maintaining a pressure above 40 mmHg is calculated. Generally, the patient 600 needs to maintain the pressure above 40 mmHg within tubing 140 for 4 to 5 seconds. When the pressure within tubing 140 is maintained above 40 mmHg for 4 to 5 seconds, the control unit 500 begins to request the cardiac rate of change of the patient 600 from the ECG monitoring unit 200. This action will cause a series of physiological responses in the cardiovascular system of the patient being tested, which are very important for the diagnosis of patent foramen ovale (PFO).

[0080] Preferably, the control unit 500 is a processor, dedicated integrated chip, or server capable of executing the information processing steps of the present invention. The control unit 500 includes a prompting unit 510, a control communication port 520, a processor 530, a memory 550, and a timer 560. The prompting unit 510, control communication port 520, memory 550, and timer 560 are respectively connected to the processor 530 via wired or wireless means to transmit information. The control communication port 520 is used for communication and is a communication port. The prompting unit 510 includes a sound reminder component and a light component. The prompting unit 510 can receive instruction information issued by the processor 530 and emit sound or light prompt information corresponding to the instruction information. The memory 550 is used to store data information of the processor 530. The timer 560 can be a timing program in the processor 530 or a specific clock. The processor 530 is used to specifically execute tasks such as data analysis and instruction generation in the control unit 500 and is the core component of the control unit 500.

[0081] In advanced medical imaging technologies, ensuring the precise injection of contrast agent 310 is crucial, affecting not only image quality but also the safety and comfort of the patient 600. Against this backdrop, the control unit 500 is designed with real-time monitoring and analysis of electrocardiogram (ECG) data in mind, and how to optimize the timing of contrast agent 310 injection based on this data. The Valsalva maneuver, a simple physiological exercise, can induce significant changes in the rate of change of the electrocardiogram, providing a unique window for monitoring and assessing cardiac function. When the patient 600 performs the Valsalva maneuver, the heart rate typically undergoes a rapid rise followed by a slow decline, with one or more inflection points indicating slowing of the heart rate. These inflection points reflect the heart's responsiveness to physiological stress and are ideal indicators for precise injection of contrast agent 310.

[0082] When the electrocardiogram rate of the patient under test 600 reaches an inflection point related to the Valsalva maneuver, the control unit 500 sends a prompt to the contrast agent injection unit 300 to administer contrast agent 310. Preferably, the inflection point refers to the point where the heart rate slows down.

[0083] For example, after receiving the electrocardiogram (ECG) data from the patient 600, the control unit 500 is responsible for analyzing the ECG data in real time. When the patient 600 performs the Valsalva maneuver, the control unit 500 analyzes the ECG rate of change in real time and identifies the inflection point of heart rate slowing through an algorithm. This judgment is based on in-depth analysis of the ECG data, including key parameters such as heart rate change trend, heart rate change amplitude, and heart rate change rate. Once the inflection point of heart rate slowing is detected, the control unit 500 immediately sends a prompt message to the contrast agent injection unit 300. This prompt message contains all necessary instructions and parameters to ensure that the contrast agent 310 is accurately injected at the optimal time. The contrast agent injection unit 300 injects immediately upon receiving the injection prompt, without allowing it to stand still or remain in place. This automation and intelligence significantly improves the efficiency and safety of contrast agent 310 use, while also reducing unnecessary repeated examinations caused by poor injection timing. This invention enables the standardization of the injection timing of contrast agent 310.

[0084] The electrocardiogram (ECG) monitoring unit 200 measures changes in heart rate, P waves, and QRS complexes. The ECG monitoring unit 200 compares the ECGs of the patient 600 before and after performing the Valsalva maneuver to measure these changes. This invention captures the cardiac physiological response induced by the Valsalva maneuver by comparing ECGs before and after the Valsalva maneuver, providing important information for subsequent diagnosis and treatment.

[0085] According to a preferred embodiment, the control unit 500 integrates an artificial intelligence semantic training model 540, which, after training, analyzes the rate of change of the electrocardiogram of the patient 600 sent by the electrocardiogram monitoring unit 200 and determines the inflection point of change related to the Valsalva maneuver.

[0086] Preferably, the artificial intelligence semantic training model 540 can be obtained by training a deep neural network (DNN). A deep neural network is a network structure composed of multiple layers of artificial neurons that simulate the workings of neurons in the human brain, enabling it to learn and recognize complex data patterns.

[0087] A key step in training the AI ​​semantic training model 540 is using a large amount of electrocardiogram (ECG) data as training samples. These samples include various patterns of normal and abnormal ECGs, especially data on the rate of change and inflection points of ECG changes related to the Valsalva maneuver. Using this data, the AI ​​semantic training model 540 learns how to identify trends in heart rate changes, particularly identifying inflection points, which is crucial for subsequent medical decision-making.

[0088] The training process of the AI ​​semantic training model 540 employs a supervised learning method, where each training sample has a corresponding label, representing the correct interpretation of the ECG data it represents (e.g., whether a change inflection point related to the Valsalva maneuver has occurred). The AI ​​semantic training model 540 gradually improves its accuracy and reliability in ECG data analysis by continuously adjusting the weights and biases in the network to minimize the difference between the predicted results and the true labels.

[0089] Furthermore, the AI ​​semantic training model 540 may employ specific techniques to optimize the training process, such as convolutional neural networks (CNNs) to extract local features from electrocardiogram (ECG) data, or recurrent neural networks (RNNs) to handle temporal dependencies in sequence data. The application of these techniques further enhances the AI ​​semantic training model 540's ability to process complex ECG data, enabling it to more accurately identify and analyze ECG rates of change and inflection points associated with Valsalva maneuvers.

[0090] While intelligent network models such as neural network models and reinforcement deep learning models can form an AI semantic training model 540 to identify inflection points related to Valsalva maneuvers, a drawback is that the intelligent network model may misjudge the inflection points related to Valsalva maneuvers that it automatically learns and identifies. Especially when the electrocardiogram (ECG) rate of change of the patient being tested differs from that of ordinary patients, the AI ​​semantic training model 540 may automatically generate inaccurate identification results for specific ECG rates, misleading doctors and leading to misdiagnosis. Therefore, based on this deficiency, the AI ​​semantic training model 540 uses the training set to retrieve training set data and identify inflection points of ECG rate changes related to Valsalva maneuvers, resulting in more accurate results. When the patient's physiological characteristics are special, or when the AI ​​semantic training model 540 cannot find approximate data from the training set, the AI ​​semantic training model 540 can issue a prompt message through the prompting unit 510 to allow medical personnel to determine the inflection points related to Valsalva maneuvers. This effectively avoids the judgment results automatically generated by the AI ​​semantic training model 540, preventing erroneous results and avoiding misleading medical staff.

[0091] Specifically, when the blowing command is issued, the control unit 500 generates labeled prompt information through the prompting unit 510. The artificial intelligence semantic training model 540 within the control unit 500 labels the electrocardiogram (ECG) change rate of the patient 600 sent by the ECG monitoring unit 200, thereby forming a training set of ECG change rates. When the patient performs the Valsalva maneuver, the ECG data will show specific change patterns, especially in heart rate. During the Valsalva maneuver, the heart rate typically increases initially and then rapidly decreases after the maneuver ends. These changes can be observed through the RR interval (the time interval between two consecutive heartbeats) of the ECG signal. Inflection point data typically refers to the start and end points of heart rate changes.

[0092] Table 1: Example data for the training set table.

[0093] In Table 1, the "@" symbol is used to connect time points and corresponding heart rate values, indicating that these data were measured at the same time point. This representation helps to clearly show heart rate data at different time points, making data reading more intuitive. For example, "70 bpm@09:05" means that at 09:05, the patient's heart rate was 70 beats per minute. This format is commonly used in medical records and data analysis to ensure that the correlation between time and related measurements is immediately apparent.

[0094] In Table 1, the initial heart rate and time represent the heart rate and time recorded before the Valsalva maneuver.

[0095] The inflection point of heart rate increase indicates the inflection point and time of heart rate rise during the Valsalva maneuver.

[0096] The heart rate decrease inflection point indicates the inflection point and time of heart rate decrease after the Valsalva maneuver ends.

[0097] Final heart rate and time represent the heart rate and time recorded after the Valsalva maneuver ends.

[0098] Heart rate variability (ΔHR / ΔT) represents the rate of change in heart rate during the Valsalva maneuver, expressed as a change in heart rate per minute.

[0099] The annotation information describes the characteristics of heart rate changes and is used to assist artificial intelligence semantic training models in understanding the data.

[0100] This invention uses an artificial intelligence semantic training model 540 to retrieve training data to analyze the rate of change in electrocardiogram and promptly determine the inflection point of heart rate change, thereby achieving a good effect of accurately reminding the contrast agent infusion time.

[0101] This invention utilizes artificial intelligence algorithms to further optimize the analysis of electrocardiogram (ECG) change rates. Through learning from a large amount of sample data, the artificial intelligence semantic training model 540 can more accurately identify ECG change features related to patent foramen ovale (PFO), thereby improving the accuracy and efficiency of diagnosis.

[0102] In response to the received start information of contrast agent intravenous infusion from the contrast agent injection unit 300, the control unit 500 issues a blowing command to the patient 600 through the prompting unit 510. And when it is determined that the blowing pressure of the patient 600 measured by the blowing unit 100 is lower than the second predetermined threshold, the control unit 500 issues a recording command of the hemodynamic status in the heart through the prompting unit 510, so that the ultrasound imaging unit 400 begins to capture images of the hemodynamic status in the heart.

[0103] To address the shortcomings of relying on work experience to determine the timing of ultrasound imaging and the inability to directly observe the patient's breathing patterns, this invention can determine the timing of ultrasound imaging based on changes in breathing patterns, thereby capturing the shunt within the heart in a timely manner and avoiding missing the optimal imaging opportunity.

[0104] According to a preferred embodiment, when the control unit 500 issues a blowing command through the prompting unit 510, the blowing unit 100 monitors the blowing pressure and its changes in the tubing 140. Based on the information condition that the blowing pressure reaches a first predetermined threshold, the control unit 500 triggers a timing command to start the connected timer 560. When the time for the blowing pressure to reach the first predetermined threshold reaches a predetermined duration, the control unit 500 generates a command message for the patient 600 to hold their breath and a timing prompt message for the duration of breath-holding. The control unit 500 of this invention can monitor changes in blowing pressure and duration through the timer 560 and the pressure sensor 110, and can also remind the patient 600 to cooperate in achieving standardized blowing pressure, thus standardizing the patient 600's Valsalva maneuver, thereby reducing the occurrence of non-standard Valsalva maneuvers by the patient 600 and improving the efficiency of the diagnostic process for patent foramen ovale.

[0105] According to a preferred embodiment, in response to the end of intravenous contrast agent infusion from the contrast agent injection unit 300, the control unit 500 issues an exhalation command to the patient 600 via the prompting unit 510, and determines whether the blowing pressure of the patient 600 measured by the blowing unit 100 is lower than a second predetermined threshold. Without a clear command, the patient 600, unsure of their breathing ability, may easily maintain a breath-holding state or inhale prematurely, which can lead to dizziness, hypoxia, and improper Valsalva maneuvering after performing the Valsalva maneuver. This invention issues clear commands by monitoring the blowing pressure, making it easier for the patient 600 to cooperate and reducing the difficulty of performing the Valsalva maneuver correctly.

[0106] For example, during the Valsalva maneuver performed by patient 600, the patient blows air through a tube at a pressure of 40 mmHg (the first predetermined threshold). After reaching this pressure, the patient holds their breath for 10 seconds (a preset duration). During an effective Valsalva maneuver, the heart rate during breath-holding will exhibit a pattern of decrease → increase → maintenance → decrease. The changes in patient 600's heart rate are compared by recording an electrocardiogram. During the maintenance phase of the heart rate (approximately 7 seconds of breath-holding), contrast agent 310 is injected. After the injection, patient 600 exhales (resumes breathing). Following exhalation, the doctor observes the intracardiac shunt using the ultrasound imaging unit 400, diagnosing the presence and grade of patent foramen ovale based on the shunt's characteristics.

[0107] This invention improves the diagnostic accuracy and operational safety of patent foramen ovale by detailing the selection and application of cardiac contrast agent 310, as well as the method for precisely setting the blowing pressure and electrocardiogram rate of change. It has important application value for clinical diagnosis.

[0108] According to a preferred embodiment, the device further includes a display unit 700. Based on changes in the blowing pressure, the control unit 500 controls the display unit 700 to display the changes in blowing pressure within different pressure ranges using a changing indicator, thereby alerting the patient 600 or medical staff to the current blowing status. This allows the patient 600 and medical staff to be aware of the current blowing status based on the changing indicator, thus enabling them to actively cooperate in completing the diagnostic process.

[0109] According to a preferred embodiment, the control unit 500 synchronously displays the ventilation pressure and its changes, timing information, and electrocardiogram changes via the display unit 700. This invention integrates these three types of information into the same display unit 700, enabling medical personnel to simultaneously monitor changes in various information types and promptly perform each operation based on received prompts, thus standardizing the diagnostic procedures for patent foramen ovale.

[0110] For example, as shown in Figure 3, the display information of the display unit 700 is divided into three lines. The top line displays red, yellow, and green lights. The middle line displays the pressure value and timing time, and the bottom line displays the electrocardiogram.

[0111] Pressure sensor 110 monitors the ventilation pressure of the patient 600 during ventilation and displays the pressure value on display unit 700. A green light illuminates when pressure is >40 mmHg, a yellow light illuminates when pressure is between 36-40 mmHg, and a red light illuminates when pressure is below 35 mmHg. No light illuminates when there is no pressure. When the pressure is >40 mmHg, timer 560 is triggered to count for 10 seconds, and the value is displayed on display unit 700.

[0112] According to a preferred embodiment, when the heart rate of the patient under test 600 slows down and reaches a preset heart rate threshold, the control unit 500 sends a prompt message to the contrast agent injection unit 300 to input contrast agent 310. Reaching the preset heart rate threshold indicates an inflection point in heart rate change. Accurate judgment and reminders at this point prevent medical staff from missing the optimal time for ultrasound imaging, thereby obtaining the heart's blood flow changes at the optimal moment, avoiding missing the best imaging opportunity, and providing the most accurate diagnostic basis.

[0113] Figures 6 and 7 show two echocardiogram images of patient A. Patient A has had contrast agent 310 injected into their body. Patient A is a 10-year-old child who, while performing the Valsalva maneuver as instructed, exhibited a resting grade of 3 and a post-Valsalva maneuver grade of 3, holding their breath for 13 seconds, and experiencing changes in heart rate. The contrast agent 310 is clearly visible and easily identifiable in the echocardiogram images. No right-to-left flow of the contrast agent 310 is observed in the echocardiogram images of Figures 6 and 7. Patient A does not have patent foramen ovale.

[0114] Figure 8 shows an echocardiogram of patient B. Patient B has received contrast agent 310. Patient B is 18 years old. During the Valsalva maneuver as instructed ("Level 1 at rest, Level 1 after Valsalva maneuver"), with a breath-holding time of 15 seconds, heart rate changes were observed. The contrast agent 310 is clearly visible and easily identifiable in the echocardiogram. In Figure 8, a flow trace of contrast agent 310 is visible from right to left in the echocardiogram. Patient B has a patent foramen ovale.

[0115] Based on experimental results of auxiliary diagnosis of patent foramen ovale based on the Valsalva maneuver, the uniform foaming treatment of contrast agent 310 in this invention allows for clear visualization of the contrast agent 310 in ultrasound images, thus providing clear diagnostic reference information. With appropriate instructions, medical personnel can promptly capture clear cardiac ultrasound images without requiring patients to repeat the Valsalva maneuver before obtaining a satisfactory image, thereby reducing the difficulty of patient cooperation.

[0116] Example 2

[0117] This embodiment is a further improvement on embodiment 1, and repeated content will not be described again.

[0118] This embodiment also provides an auxiliary diagnostic method for patent foramen ovale based on the Valsalva maneuver, as shown in Figure 5, which includes the following steps.

[0119] S000: During the preparation of contrast agent 310, contrast agent 310 is uniformly mixed with air by repeatedly pulling the syringe 320 to form a uniform mixture of air and contrast agent 310.

[0120] S100: When the blowing pressure of the patient 600, as fed back by the blowing unit 100, reaches a first predetermined threshold and remains there for a predetermined duration, the control unit 500 begins to request the electrocardiogram (ECG) rate of change of the patient 600 output by the ECG monitoring unit 200. This invention utilizes the control unit 500 and the pressure sensor 110 to collect data on the blowing pressure and duration, thereby providing an accurate timing for judging ECG changes and preventing the patient 600 from repeatedly performing the Valsalva maneuver.

[0121] S110: When the control unit 500 issues a blowing command through the prompting unit 510, the blowing unit 100 monitors the blowing pressure of the tubing 140 and its changes. The control unit 500 triggers a timing command to start the timer 560 connected to it based on the information condition that the blowing pressure reaches a first predetermined threshold. When the time for the blowing pressure to reach the first predetermined threshold reaches a predetermined duration, the control unit 500 generates a command message for the patient 600 to hold their breath and a timing prompt message for the duration of breath-holding.

[0122] S200: When the electrocardiogram rate of the patient under test 600 is determined to have an inflection point related to the Valsalva maneuver, the control unit 500 sends a prompt to the contrast agent injection unit 300 to input contrast agent 310. This invention analyzes the electrocardiogram rate of change and determines the timing of the inflection point, enabling the contrast agent injection unit 300 to promptly inject contrast agent 310 intravenously, avoiding missed injection opportunities and standardizing the injection timing of contrast agent 310.

[0123] S210: When the heart rate of the patient under test (600) slows down and reaches a preset heart rate threshold, the control unit 500 sends a prompt to the contrast agent injection unit 300 to input contrast agent 310. Reaching the preset heart rate threshold indicates an inflection point in heart rate change. Accurate judgment and alerts at this point prevent medical staff from missing the optimal time for ultrasound imaging, thus obtaining the best possible information on cardiac blood flow changes and providing the most accurate diagnostic information. The preset heart rate threshold is, for example, 5-10 beats / min.

[0124] S220: Infuse 1 ml of air + 9 ml of normal saline (contrast agent 310) into a vein.

[0125] S300: In response to the received start information of contrast agent intravenous infusion from the contrast agent injection unit 300, the control unit 500 issues a breathing command to the patient 600 via the prompting unit 510. Furthermore, if the breathing pressure of the patient 600 measured by the breathing unit 100 is determined to be lower than a second predetermined threshold, the control unit 500 issues a recording command for the hemodynamic status within the heart via the prompting unit 510. Addressing the shortcomings of relying on work experience to determine the timing of imaging and the inability to directly observe the breathing status of the patient 600, this invention can determine the timing of ultrasound imaging based on changes in the breathing status, thereby capturing the shunt situation within the heart in a timely manner and avoiding missing the optimal imaging opportunity.

[0126] S310: In response to the end information of contrast agent intravenous infusion of the contrast agent injection unit 300, the control unit 500 issues an exhalation command to the patient 600 through the prompting unit 510, and determines whether the exhalation pressure of the patient 600 measured by the exhalation unit 100 is lower than the second predetermined threshold.

[0127] S320: The control unit 500 sends a reminder command to the ultrasound imaging unit 400 to image the hemodynamic conditions within the heart.

[0128] Example 3

[0129] This embodiment is a further elaboration of Embodiments 1 and 2, and repeated content will not be repeated.

[0130] This embodiment provides a processor for constructing an artificial intelligence dataset related to cardiac contrast agent injection. As shown in Figure 9, the processor 530 includes an artificial intelligence semantic training model 540. When the processor 530 issues a blowing command to the patient 600 via the prompting unit 510, the processor 530 receives the blowing pressure of the patient 600 from the blowing unit 100 and the electrocardiogram (ECG) changes of the patient 600 from the ECG monitoring unit 200. When the prompting unit 510 generates labeled prompt information, the artificial intelligence semantic training model 540 within the processor 530 labels the ECG change rate of the patient 600 sent by the ECG monitoring unit 200, preparing for the construction of a training set for the ECG change rate. The artificial intelligence semantic training model 540 retrieves the training set to analyze the ECG change rate and promptly determines the inflection point of the ECG change rate, thereby accurately reminding the patient of the contrast agent 310 input time.

[0131] In order for the AI ​​semantic training model 540 to learn to identify the inflection point of the electrocardiogram rate of change, it is necessary to accurately label various types of data during the data collection process and express relevant information at appropriate time points.

[0132] When the patient begins the Valsalva maneuver, the ECG monitoring unit 200 is immediately activated, initiating real-time monitoring of the rate of change in cardiac electrical activity. The processor 530 receives the ECG signal from the monitoring unit 200 and performs preliminary analysis using the artificial intelligence semantic training model 540. This process begins at the time of the start of the breath (T0). start Begin by recording initial data on the rate of change of the electrocardiogram.

[0133] For example, electrodes are attached to the patient's body surface to acquire electrocardiogram (ECG) signals. The ECG monitoring unit 200 acquires the raw ECG signals from the electrodes. The ECG monitoring unit 200 transmits the ECG signals to the processor 530 via a wired connection or wireless means (such as Bluetooth or Wi-Fi). The processor 530 preprocesses the received signals to remove interference and noise. The processor 530 converts the ECG signals from analog to digital form for subsequent analysis. The processor 530 uses an artificial intelligence semantic training model 540 to extract key features from the ECG signals, such as the P wave and QRS complex. The rate of change of the ECG is calculated from the T wave... start Start recording initial data.

[0134] For example, when the processor 530 is a microcontroller (MCU), the GPIO (General Purpose Input / Output) pins are the basic interface of the microcontroller, used for reading and writing digital signals. The ECG monitoring unit 200 is connected to the GPIO pins. The ADC module is connected to the GPIO pins to convert analog signals into digital signals for microcontroller processing. SPI is connected to the ADC module for high-speed communication between the microcontroller and peripherals.

[0135] First, the start signal for the blowing action is read via the GPIO pin, initiating monitoring. Then, the analog ECG signal is received from the ECG monitoring unit 200 using the ADC (Analog-to-Digital Converter) module and converted into a digital signal. The ECG signal data is transmitted to the MCU via the Serial Peripheral Interface (SPI) or I2C protocol, where noise filtering and signal preprocessing are performed using the MCU's internal DSP library. Next, the MCU extracts features and calculates the rate of change using a pre-trained model stored in Flash memory, and monitors feature changes in real time via an interrupt mechanism, generating alarms. The processed data is stored via an external EEPROM or SD card module, generating a detailed report, and the data is transmitted to an external system for further analysis via a UART or USB interface.

[0136] For example, when the processor 530 is a digital signal processor (DSP), an external trigger signal is input to the DSP's trigger input port. The analog output pin of the ECG monitoring unit 200 is connected to the input pin of the ADC module. The digital output port of the ADC module is connected to the input port of the DSP's high-speed parallel interface (EMIF). DSP monitoring is first started by an external trigger signal. The high-speed ADC module acquires signals from the ECG monitoring unit 200, and then transmits the signal data to the DSP memory through a high-speed parallel interface (such as EMIF). Signal preprocessing is performed using the DSP's built-in filtering and Fourier transform libraries, and complex feature extraction and rate of change calculation are performed using the DSP's high-performance computing capabilities.

[0137] For example, when the processor 530 is a single-board computer, the startup signal source is connected to the GPIO input pin of the single-board computer. The output pin of the ECG monitoring unit 200 is connected to the I2C or SPI interface input pin of the single-board computer. A Python script and GPIO pins are used to monitor the startup signal. The single-board computer connects to the ECG monitoring unit 200 via the I2C or SPI interface to acquire signal data and uses memory mapping technology to transfer the data to the Raspberry Pi's processor. Signal preprocessing uses the NumPy and SciPy libraries for noise filtering and signal processing, and runs a pre-trained machine learning model (such as TensorFlow or PyTorch) for feature extraction and rate of change calculation.

[0138] When the blowing pressure reaches a first predetermined threshold (e.g., 40 mmHg), the blowing unit 100 detects this and immediately transmits the information to the processor 530. At this time, the processor 530 records the time point (T) when the blowing pressure reaches the predetermined threshold. threshold ), and switch to a more intensive monitoring mode, focusing on the magnitude of sudden changes in the rate of change of the electrocardiogram. Recording begins with exhalation (T... start ) until the blowing pressure reaches the standard (T) threshold ECG variability data during the period.

[0139] Processor 530 analysis from T start To T threshold During the monitoring period, the electrocardiogram (ECG) rate of change data was used to identify abrupt changes. When the detected abrupt change exceeded the maximum change during the previous monitoring period, and the extent of the exceedance reached or exceeded a preset percentage (e.g., 30%, 50%, 70%), the time point (T0) at which the abrupt change occurred was recorded. inflection This was identified as the inflection point in the rate of change of the electrocardiogram that was related to the Valsalva maneuver.

[0140] In determining the inflection point (T) of the rate of change of electrocardiographic change. inflectionAfterwards, processor 530 immediately sends a prompt message to contrast agent injection unit 300 indicating preparation for injection. Medical personnel, following the instructions of prompt unit 510, prepare to inject the contrast agent into the vein of the patient to be tested. The time point (T) when the prompt message is sent is recorded. prepare ).

[0141] After confirming that the electrocardiogram (ECG) rate of change meets preset conditions, the processor 530 sends an injection start signal to the contrast agent injection unit 300. While medical personnel perform the injection, the processor 530 continues to monitor the ECG rate of change in real time to ensure that the injection timing coincides with the inflection point (Tc) of the ECG rate of change. inflection )synchronous.

[0142] After injection, the processor 530 sends an exhalation command to the patient 600 via the prompting unit 510 and instructs the ultrasound imaging unit 400 to begin recording the dynamic blood flow within the heart. The time point at which the exhalation command is sent is recorded (T0). exhale ) and the time point at which the ultrasound imaging unit is activated to capture and record changes in cardiac blood flow (T imaging ).

[0143] The format for labeling the electrocardiogram change rate of the patient 600 sent by the electrocardiogram monitoring unit 200 by the artificial intelligence semantic training model 540 is shown in Table 2.

[0144] [Revised according to Rule 26, 11.05.2026] Table 2: Example of labeling information for ECG rate of change

[0145] Through the detailed time point annotations and information descriptions described above, the artificial intelligence semantic training model 540 can learn and identify the inflection point of the electrocardiogram rate of change, and can accurately determine the optimal time for contrast agent 310 input.

[0146] The process by which the AI ​​semantic training model 540 learns to determine the inflection point of the rate of change in electrocardiogram is shown below.

[0147] [Revised according to Rule 26, 2026, November 5, 2026] Normalization processing of ECG signals was performed:

[0148] Where X(t) is the original electrocardiogram signal, μ is the mean, and σ is the standard deviation.

[0149] [Revised according to Rule 26, 11.05.2026] Extract features from the normalized signal, such as the RR interval and P wave. Assume the RR interval is used as the primary feature: RR... i =R i -R i-1 (2).

[0150] Among them, R i This represents the time point of the i-th R-wave.

[0151] [Revised according to Rule 26, 11.05.2026] A sliding window is used to calculate the abrupt change in the rate of change of electrocardiogram (ECG) variability. Assuming the window size is w, the rate of change of the data within the window is:

[0152] [Revised according to Rule 26, 11.05.2026] Mutation detection can be performed using the difference method. The difference method is: d(t)=RR(t)-RR(t-1) (4).

[0153] When |d(t)| is greater than a certain threshold δ, it is determined to be an inflection point. If |d(t)|>δ, then a change point is detected at time t.

[0154] The feature sequence is trained using a Long Short-Term Memory (LSTM) network. Assume the input features are X = [RR1, RR2, ..., RR...]. n The AI ​​semantic training model 540 outputs a prediction.

[0155] [Revised according to Rule 26, 2026] The loss function of the AI ​​semantic training model 540 is defined as the mean squared error:

[0156] Optimize the parameters θ of AI semantic training model 540 using gradient descent:

[0157] [Revised according to Rule 26, 2026] Where η is the learning rate.

[0158] [Revised according to Detailed Rule 26, 2026] The inflection point of the electrocardiogram rate of change predicted by the AI ​​semantic training model 540 is:

[0159] When the predicted value When the value exceeds the first preset threshold α, it is determined to be the inflection point of the electrocardiogram rate of change, that is, if... The inflection point of the rate of change in electrocardiogram was detected at time t.

[0160] [Revised according to Rule 26, 11.05.2026] In practical applications, the AI ​​semantic training model 540 monitors ECG data in real time, identifies inflection points of change, and determines the optimal injection timing using the following formula: Set the injection timing threshold β, then...

[0161] The specific timing of the injection is calculated as follows:

[0162] [Revised according to Rule 26, 11.05.2026] If an inflection point in the rate of change of electrocardiogram is detected at time t and the injection conditions are met: t 注射 =t(9).

[0163] The process by which the artificial intelligence semantic training model 540 learns to determine the timing of contrast agent 310 injection based on the labeled information of the blowing pressure is shown below.

[0164] [Revised according to Rule 26, 11.05.2026] The received air pressure data is smoothed:

[0165] P(t) represents pressure data, and t represents time. smoothed (t) represents the pressure data after smoothing at time t, and w is the width of the smoothing window.

[0166] [Revised according to Rule 26, 11.05.2026] The normalization process is as follows: map the pressure data to the range [0,1].

[0167] P norm (t) represents the pressure data after normalization at time t, P min and P max These represent the minimum and maximum values ​​of the pressure data, respectively.

[0168] [Revised according to Rule 26, 2026, November 5th] The formula for calculating the rate of change of blowing pressure is:

[0169] ΔP(t) represents the rate of change of the blowing pressure at time t, and Δt represents the time step.

[0170] [Revised according to Rule 26, 11.05.2026] The time when the blowing pressure reaches the second predetermined threshold is: T pressure_low =min{t|P norm (t)≤P threshold} (13).

[0171] T pressure_low This indicates the time it takes for the blowing pressure to drop to the second set threshold. (P) threshold This indicates the second predetermined pressure threshold.

[0172] [Revised according to Rule 26, 2026, 11.05.2026] The formula for the AI ​​semantic training model 540 to predict the optimal shooting time using air pressure and injection information is as follows:

[0173] is the predicted imaging alert instruction time, and f(·) is the prediction function.

[0174] [Revised according to Rule 26, 11.05.2026] The formula for issuing an imaging alert command after the artificial intelligence semantic training model 540 confirms that the blowing pressure has reached the second predetermined threshold is: if P norm (t)≤P threshold ,then T maging_remind =t (15).

[0175] [Revised according to Rule 26, 11.05.2026] Use an untrained dataset to validate the model's accuracy and calculate the prediction error:

[0176] Where M is the number of validation samples. It is the prediction time for the i-th sample. That is the actual time.

[0177] [Revised according to Rule 26, 2026, November 5th] The formula, further optimized based on feedback from medical staff, is as follows:

[0178] θ new These are the updated model parameters, θ old These are the model parameters before the update, Δθ feedback These parameters are adjusted based on feedback information.

Claims

1. A method for constructing an artificial intelligence dataset related to cardiac contrast agent injection, the method comprising: The blowing pressure of the patient (600) is measured by the blowing unit (100); The electrocardiogram (ECG) changes of the patient (600) are measured by the ECG monitoring unit (200); The method is characterized in that it further includes: The control unit (500) determines the injection timing of the contrast agent injection unit (300) and issues a recording command for the hemodynamic conditions within the heart; When the blowing pressure reaches the first predetermined threshold, the control unit (500) monitors the sudden change in the electrocardiogram rate of change in real time; when the sudden change exceeds the change from the time point when blowing begins to the time point when the blowing pressure reaches the first predetermined threshold, the time point when the sudden change occurs is determined as the inflection point of the electrocardiogram rate of change. After determining the inflection point of the electrocardiogram rate of change, the control unit (500) sends a prompt message to the contrast agent injection unit (300) indicating that injection is ready. After confirming that the electrocardiogram rate of change meets the preset conditions, the control unit (500) sends an injection start signal to the contrast agent injection unit (300); The artificial intelligence semantic training model (540) in the control unit (500) labels the electrocardiogram change rate of the patient (600) sent by the electrocardiogram monitoring unit (200), thereby forming a training set of electrocardiogram change rates.

2. The method for constructing an artificial intelligence dataset according to claim 1, characterized in that, The method further includes: The control unit (500) analyzes the electrocardiogram rate of change data from the start of exhalation to the attainment of target pressure, and looks for abrupt changes. Once the detected mutation amplitude exceeds the maximum change amplitude during the previous monitoring period, and the degree of excess reaches or exceeds a preset percentage, the time point at which the mutation amplitude occurs is determined as the inflection point of the ECG rate of change related to the Valsalva maneuver.

3. The method for constructing an artificial intelligence dataset according to claim 1 or 2, characterized in that, The method further includes: when the blowing pressure of the patient (600) being tested, as fed back by the blowing unit (100), reaches a first predetermined threshold and remains there for a predetermined duration, the control unit (500) begins to request the electrocardiogram change rate of the patient (600) being tested, output by the electrocardiogram monitoring unit (200).

4. The method for constructing an artificial intelligence dataset according to any one of claims 1 to 3, characterized in that, The method further includes: In response to the received start information of contrast agent intravenous input from the contrast agent injection unit (300), the control unit (500) issues a blowing command to the patient to be tested (600) through the prompting unit (510), and the blowing unit (100) monitors the blowing pressure of the tubing (140) and its changes; The control unit (500) triggers a timing command to start the timer (560) connected to it based on the information condition that the blowing pressure reaches a first predetermined threshold. When the time for the blowing pressure to reach the first predetermined threshold reaches a predetermined duration, the control unit (500) generates a command message to make the patient (600) hold their breath and a timing prompt message for the duration of breath-holding.

5. The method for constructing an artificial intelligence dataset according to any one of claims 1 to 4, characterized in that, The method for confirming that the electrocardiogram rate of change meets the preset conditions includes: When the heart rate of the patient to be tested (600) slows down and reaches a preset heart rate threshold, the control unit (500) sends a prompt message to the contrast agent injection unit (300) to input the contrast agent (310).

6. The method for constructing an artificial intelligence dataset according to any one of claims 1 to 5, characterized in that, The method further includes: After the injection is completed, the control unit (500) sends an exhalation command to the patient (600) through the prompting unit (510) and instructs the ultrasound imaging unit (400) to start recording the blood flow dynamics in the heart; The control unit (500) responds to the end information of the contrast agent intravenous infusion from the contrast agent injection unit (300). The prompting unit (510) issues an exhalation command to the patient (600) to be tested and determines whether the exhalation pressure of the patient (600) measured by the exhalation unit (100) is lower than a second predetermined threshold. When the blowing pressure is lower than the second predetermined threshold, the control unit (500) issues a recording instruction for the hemodynamic status of the heart via the prompting unit (510).

7. An apparatus for constructing an artificial intelligence dataset related to cardiac contrast agent injection, the apparatus comprising: The blowing unit (100) is used to measure the blowing pressure of the patient (600) to be tested; An electrocardiogram monitoring unit (200) is used to measure the electrocardiogram changes of the patient (600) to be tested; A contrast agent injection unit (300) is used to administer contrast agent (310) to the patient (600) to be tested; Its characteristic is that it further includes: Control unit (500): Used to determine the injection timing of the contrast agent injection unit (300) and to issue recording instructions for the hemodynamic conditions within the heart; The control unit (500) is configured as follows: When the blowing pressure reaches the first predetermined threshold, the control unit (500) monitors the sudden change in the electrocardiogram rate of change in real time; when the sudden change exceeds the change range from the start of blowing to the time when the blowing pressure reaches the first predetermined threshold, the time point where the sudden change occurs is determined as the inflection point of the electrocardiogram rate of change; after determining the inflection point of the electrocardiogram rate of change, the control unit (500) sends a prompt message to the contrast agent injection unit (300) indicating that injection is ready; After confirming that the electrocardiogram rate of change meets the preset conditions, the control unit (500) sends an injection start signal to the contrast agent injection unit (300); The artificial intelligence semantic training model (540) in the control unit (500) labels the electrocardiogram change rate of the patient (600) sent by the electrocardiogram monitoring unit (200), thereby forming a training set of electrocardiogram change rates.

8. The artificial intelligence dataset construction apparatus according to claim 7, characterized in that, The control unit (500) analyzes the electrocardiogram rate of change data from the start of exhalation to the attainment of target pressure, and looks for abrupt changes. Once the detected mutation amplitude exceeds the maximum change amplitude during the previous monitoring period, and the degree of excess reaches or exceeds a preset percentage, the time point at which the mutation amplitude occurs is determined as the inflection point of the ECG rate of change related to the Valsalva maneuver.

9. The artificial intelligence dataset construction apparatus according to claim 7 or 8, characterized in that, After the injection is completed, the control unit (500) sends an exhalation command to the patient (600) through the prompting unit (510) and instructs the ultrasound imaging unit (400) to start recording the blood flow dynamics in the heart; In response to the end information of the contrast agent intravenous infusion of the contrast agent injection unit (300), the control unit (500) sends an exhalation command to the patient to be tested (600) through the prompting unit (510) and determines whether the exhalation pressure of the patient to be tested (600) measured by the exhalation unit (100) is lower than a second predetermined threshold. When the blowing pressure is lower than the second predetermined threshold, the control unit (500) issues a recording instruction for the hemodynamic status of the heart via the prompting unit (510).

10. The artificial intelligence dataset construction apparatus according to any one of claims 7 to 9, characterized in that, When the blowing pressure of the patient (600) being tested, as fed back by the blowing unit (100), reaches a first predetermined threshold and remains there for a predetermined duration, the control unit (500) begins to request the rate of change of the electrocardiogram of the patient (600) being tested, output by the electrocardiogram monitoring unit (200).

11. The artificial intelligence dataset construction apparatus according to any one of claims 7 to 10, characterized in that, In response to the received start information of contrast agent intravenous input from the contrast agent injection unit (300), the control unit (500) issues a blowing command to the patient to be tested (600) through the prompting unit (510), and the blowing unit (100) monitors the blowing pressure of the tubing (140) and its changes; The control unit (500) triggers a timing command to start the timer (560) connected to it based on the information condition that the blowing pressure reaches a first predetermined threshold; if the time for the blowing pressure to reach the first predetermined threshold reaches a predetermined duration, the control unit (500)... The control unit (500) generates instructions to make the patient (600) hold their breath and timing information for the duration of breath-holding.

12. A diagnostic device for patent foramen ovale based on the Valsalva maneuver, characterized in that, The blowing unit (100) is used to measure the blowing pressure of the patient (600) to be tested; An electrocardiogram monitoring unit (200) is used to measure the electrocardiogram changes of the patient (600) to be tested; A contrast agent injection unit (300) is used to administer contrast agent (310) to the patient (600) to be tested; Its features are, When the blowing pressure of the patient (600) being tested, as fed back by the blowing unit (100), reaches a first predetermined threshold and remains there for a predetermined duration, the control unit (500) begins to request the rate of change of the electrocardiogram of the patient (600) being tested, output by the electrocardiogram monitoring unit (200). When it is determined that the electrocardiogram rate of change of the patient to be tested (600) has a change inflection point related to the Valsalva maneuver, the control unit (500) sends a prompt message to the contrast agent injection unit (300) to input the contrast agent (310), and synchronizes the timing of contrast agent injection with the change inflection point of electrocardiogram rate of change. In response to the received start information of contrast agent intravenous input from the contrast agent injection unit (300), the control unit (500) issues a blowing command to the patient to be tested (600) through the prompting unit (510), and when it is determined that the blowing pressure of the patient to be tested (600) measured by the blowing unit (100) is lower than a second predetermined threshold, the control unit (500) issues a recording command of the hemodynamic status of the heart through the prompting unit (510) to record the dynamics of cardiac blood flow; The determination of the inflection point of the rate of change of the electrocardiogram associated with the Valsalva maneuver in time is performed as follows: When the mutation amplitude exceeds the change amplitude from the time point when the exhalation begins to the time point before the exhalation pressure reaches the first predetermined threshold, the time point when the mutation amplitude occurs is determined as the inflection point; the control unit (500) analyzes the electrocardiogram change rate data from the start of exhalation to the time point when the pressure reaches the target, and finds the mutation amplitude; when the mutation amplitude is detected to exceed the maximum change amplitude during the previous monitoring period, and the degree of exceedance reaches or exceeds a preset percentage, the time point when the mutation amplitude occurs is determined as the inflection point of the electrocardiogram change rate related to the Valsalva maneuver.

13. The auxiliary diagnostic device for patent foramen ovale based on Valsalva maneuver according to claim 12, characterized in that, When a blowing command is issued, the control unit (500) generates a labeled prompt information through the prompting unit (510). The artificial intelligence semantic training model (540) in the control unit (500) labels the electrocardiogram change rate of the patient (600) sent by the electrocardiogram monitoring unit (200), thereby forming a training set of electrocardiogram change rate.

14. The auxiliary diagnostic device for patent foramen ovale based on Valsalva maneuver according to claim 12 or 13, characterized in that, When the control unit (500) issues a blowing command through the prompting unit (510), the blowing unit (100) monitors the blowing pressure and its changes in the pipeline (140). The control unit (500) triggers a timing command to start the timer (560) connected to it based on the information condition that the blowing pressure reaches a first predetermined threshold. When the time for the blowing pressure to reach the first predetermined threshold reaches a predetermined duration, the control unit (500) generates instruction information for the patient to be tested (600) to hold their breath and timing information for the duration of breath-holding.

15. The auxiliary diagnostic device for patent foramen ovale based on Valsalva maneuver according to any one of claims 12 to 14, characterized in that, The air blowing unit (100) includes a pressure sensor (110), a pressure communication port (120), an air blowing assembly (130), and a pipe (140). The pressure sensor (110) is disposed in the pipe (140) connected to the air blowing assembly (130) to collect the air blowing pressure; The pressure sensor (110) sends the blowing pressure to the control unit (500) through the pressure communication port (120). Data information.

Images