Postoperative thrombus prediction device and method, ventricular assist system, storage medium and equipment

Abstract

The invention relates to a postoperative thrombus prediction device and method, a ventricular assist system, a storage medium and equipment.The device comprises a data collection module, a calculation module and a thrombus prediction module.The data collection module is configured to collect equipment operation parameters, physiological parameters and anticoagulation parameters in the operation process of an interventional ventricular assist device; the calculation module is configured to calculate a multi-dimensional evaluation index based on the equipment operation parameters, the physiological parameters and the anticoagulation parameters, and the thrombus prediction module is configured to estimate the thrombus formation probability through a preset thrombus prediction model according to the multi-dimensional evaluation index.

Description

Technical Field

[0001] The present invention relates to the technical field of ventricular assist devices, and in particular to a postoperative thrombus prediction device, method, ventricular assist system, storage medium and equipment. Background Art

[0002] With the widespread use of interventional ventricular assist devices (VADs) in the treatment of patients with severe heart failure, thrombosis has become a significant complication affecting patient prognosis. VADs alter blood flow patterns within the heart and major vessels through the VAD system, potentially leading to slowed or stagnant blood flow in certain areas. Furthermore, the high-speed operation of the catheter pump, which uses an internal motor, increases blood shear stress, damaging red blood cells and activating platelets, significantly increasing the risk of thrombosis.

[0003] In current clinical practice, medical staff usually assess the risk of thrombosis by closely monitoring the patient's coagulation function after surgery. However, the existing monitoring methods have obvious lags. In addition, conventional monitoring cannot capture the dynamic parameter changes during the operation of the ventricular assist device, making it difficult to reflect the early signs of thrombosis in a timely manner.

[0004] Therefore, how to predict the risk of thrombosis during the operation of ventricular assist devices has become a technical problem that urgently needs to be solved. Summary of the Invention

[0005] In view of this, embodiments of the present invention provide a postoperative thrombus prediction device, method, ventricular assist system, storage medium and equipment to solve the problems existing in the prior art.

[0006] The present invention adopts the following technical solutions:

[0007] In a first aspect, an embodiment of the present invention provides a postoperative thrombosis prediction device, comprising:

[0008] a data acquisition module configured to acquire device operating parameters, physiological parameters, and anticoagulation parameters during operation of the interventional ventricular assist device;

[0009] a calculation module configured to calculate a multidimensional evaluation index based on device operating parameters, physiological parameters, and anticoagulation parameters;

[0010] The thrombosis prediction module is configured to estimate the probability of thrombosis formation through a preset thrombosis prediction model based on the multidimensional evaluation index.

[0011] In a second aspect, an embodiment of the present invention provides a method for predicting postoperative thrombosis, comprising:

[0012] Collect device operating parameters, physiological parameters and anticoagulation parameters during the operation of interventional ventricular assist devices;

[0013] Calculate a multidimensional assessment index based on device operating parameters, physiological parameters, and anticoagulation parameters;

[0014] Based on the multidimensional assessment index, the probability of thrombosis is estimated by the preset thrombosis prediction model.

[0015] In a third aspect, an embodiment of the present invention provides a ventricular assist system comprising the postoperative thrombus prediction device as described above.

[0016] In a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium, wherein the storage medium stores at least one instruction, and the at least one instruction is loaded by a processor and executed to implement the postoperative thrombosis prediction method as described above.

[0017] In a fifth aspect, an embodiment of the present invention provides an electronic device, comprising a processor and a memory, wherein the memory stores at least one instruction, and the instruction is loaded and executed by the processor to implement the postoperative thrombosis prediction method as described above.

[0018] One embodiment of the above invention has the following advantages or beneficial effects:

[0019] An embodiment of the present invention mainly provides a postoperative thrombosis prediction device, which integrates the equipment operating parameters, physiological parameters and anticoagulation parameters of the interventional ventricular assist device, and can calculate the two core evaluation indicators, the mechanical-physiological coupling index and the blood stasis index, respectively, and use a preset thrombosis prediction model to estimate the probability of postoperative thrombosis formation, so that the risk of thrombosis can be quantified during the ventricular assist process.

[0020] Compared with traditional thrombosis risk assessment schemes, the technical solution of the present invention breaks through the limitations of single parameter monitoring, improves the accuracy and reliability of thrombosis prediction, provides important guarantees for the safe operation of interventional ventricular assist devices, and effectively reduces the risk of postoperative thrombotic complications. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments, which constitute a part of the present invention. The exemplary embodiments of the present invention and their descriptions are provided to explain the present invention and do not constitute an undue limitation of the present invention. In the drawings:

[0022] Figure 1 This is a structural block diagram of a postoperative thrombosis prediction device provided by one embodiment of the present invention;

[0023] Figure 2 This is a flowchart of a postoperative thrombosis prediction method provided by one embodiment of the present invention. DETAILED DESCRIPTION

[0024] To make the objectives, technical solutions, and advantages of the present invention more clear, the technical solutions of the present invention will be clearly and completely described below in conjunction with specific embodiments of the present invention and corresponding drawings. In the description of the present invention, it should be noted that the term "or" is generally used in the sense of including "and / or" unless the content clearly indicates otherwise.

[0025] In the description of the present invention, it should be noted that, unless otherwise expressly specified or limited, the terms "mounted," "connected," and "connected" should be understood in a broad sense. In addition, in the description of this application, the terms "first," "second," etc. are used only to distinguish descriptions and should not be understood to indicate or imply relative importance.

[0026] Obviously, the embodiments described are only some of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0027] In order to solve the problems existing in the prior art, the present invention provides a postoperative thrombosis prediction device in a preferred embodiment, such as Figure 1 The device includes at least a data acquisition module 11, a calculation module 12 and a thrombus prediction module 13, wherein the data acquisition module 11 is configured to collect device operating parameters, physiological parameters and anticoagulation parameters during the operation of the interventional ventricular assist device, the calculation module 12 is configured to calculate a multidimensional evaluation index based on the device operating parameters, physiological parameters and anticoagulation parameters, and the thrombus prediction module 13 is configured to estimate the probability of thrombosis through a preset thrombus prediction model based on the multidimensional evaluation index.

[0028] In a preferred embodiment, the device operating parameters collected by the data acquisition module 11 include the motor speed, current and catheter pump flow of the interventional ventricular assist device; physiological parameters include left ventricular pressure and aortic pressure; and anticoagulation parameters include the cumulative injection volume of heparin.

[0029] Preferably, an invasive ventricular assist device consists of a catheter pump implanted in the patient's body and connected to an external control device. The catheter pump can be inserted into the transvalvular position to increase blood flow, reduce myocardial oxygen consumption, and help enhance cardiac pumping function in patients with acute heart failure in the short term. The external control device is responsible for controlling the operating parameters of the catheter pump, monitoring system status, and recording various data during treatment to support clinical decision-making.

[0030] Preferably, the catheter pump is integrated with a drive motor and a pressure sensor, wherein the drive motor is located at the proximal part of the catheter pump and drives the pump blades to operate through high-speed rotation, thereby drawing blood from the left ventricle and delivering it to the aorta to achieve an auxiliary pumping function for the left ventricle; the pressure sensors are respectively arranged at the distal and proximal ends of the catheter pump, the pressure sensor located at the distal end is used to obtain left ventricular pressure data, and the sensor located at the proximal end is used to obtain aortic pressure data.

[0031] Preferably, the speed signal and current signal of the driving motor can be monitored in real time by an extracorporeal control device, wherein the speed signal can reflect the working status of the catheter pump, and the current signal not only provides motor power consumption information, but also indirectly reflects the change in resistance in the pump, helping to identify possible thrombosis or mechanical failure.

[0032] Optionally, the flow rate of the catheter pump may be directly measured by a flow sensor, or calculated and estimated based on a preset speed-current-flow mathematical model, which is not specifically limited in this embodiment.

[0033] In a preferred embodiment, the cumulative heparin injection volume is obtained in real time through the data interface between the extracorporeal control device and the infusion system. The interface is connected to the heparin infusion pump used during the operation, and the dose and time of each heparin injection are recorded. The data acquisition module 11 automatically accumulates and calculates the total injection volume (unit IU) and records these data synchronously with other parameters.

[0034] In a preferred embodiment, the data acquisition module 11 is further configured to preprocess the collected raw data to ensure the accuracy of subsequent calculations. Preferably, preprocessing includes at least data synchronization, data denoising, and outlier detection. By preprocessing various parameters, the reliability of subsequent thrombosis risk assessment can be effectively improved.

[0035] Preferably, the data acquisition module 11 uses a synchronous acquisition mechanism to ensure the time consistency of the device operating parameters and physiological parameters. Preferably, the data acquisition frequency is set to once every 30ms, and each record contains a millisecond-accurate timestamp (YYYY-MM-DD HH:MM:SS.sss), as well as the device operating parameters and physiological parameters at the corresponding moment.

[0036] Preferably, the data acquisition module 11 also performs data structuring in the preprocessing stage, organizing the synchronously collected parameters into a structured columnar data format, where each row of data corresponds to a sampling time point, including the timestamp and all parameter values ​​of the time point, so that when the calculation module 12 subsequently efficiently processes a large amount of time series data, it can quickly extract data from a specific time window for analysis.

[0037] In a preferred embodiment, for left ventricular pressure and aortic pressure, the data acquisition module 11 uses a second-order Butterworth low-pass filter for processing, and sets the cutoff frequency to 5 Hz to remove high-frequency noise such as electromagnetic interference and catheter jitter to ensure the data quality of subsequent analysis.

[0038] In a preferred embodiment, the data acquisition module 11 also includes an outlier detection function. By setting reasonable ranges and rate-of-change ranges for various parameters, it identifies and marks possible outliers. For data points that significantly deviate from the physiologically acceptable range, the data acquisition module 11 marks them as invalid data and excludes them from subsequent calculations to prevent outliers from interfering with the evaluation results. For data that deviates slightly but remains within an acceptable range, the data acquisition module 11 uses a local smoothing algorithm for correction. In this embodiment, the corresponding correction algorithm is not specifically defined, and those skilled in the art may adapt it to their needs.

[0039] In a preferred embodiment, the multidimensional evaluation index calculated by the calculation module 12 includes a mechanical-physiological coupling index and a blood stasis index, wherein the calculation module 12 can determine the mechanical-physiological coupling index based on the device operating parameters, physiological parameters and anticoagulation parameters; the calculation module 12 can determine the blood stasis index based on the device operating parameters and physiological parameters.

[0040] In a preferred embodiment, the calculation module 12 is further configured to: respectively determine the product of the motor speed and the catheter pump flow, the pressure difference between the aortic pressure and the left ventricular pressure, and the logarithmic value of the cumulative heparin injection volume; and determine a mechanical-physiological coupling index based on the correlation between the product, the pressure difference, and the logarithmic value; wherein the mechanical-physiological coupling index is positively correlated with the product and the logarithmic value, and negatively correlated with the pressure difference.

[0041] The mechanical-physiological coupling index (MPCI) aims to quantify the dynamic interaction between the catheter pump operating state and the patient's physiological state, while taking into account the impact of anticoagulation therapy. Its specific calculation formula is as follows:

[0042]

[0043] Specifically, the product of motor speed and catheter pump flow rate reflects the mechanical output power and operating status of the catheter pump. High speed and high flow rate indicate that the catheter pump maintains adequate hemodynamics, effectively reducing the risk of thrombosis caused by blood retention. High speed and / or low flow rate decrease the product of the two, increasing the risk of thrombosis.

[0044] The pressure difference between aortic pressure and left ventricular pressure, known as the transvalvular pressure gradient, reflects the resistance the heart must overcome. A higher transvalvular pressure gradient means the heart requires greater force to pump blood. If the catheter pump is inefficient, this can lead to unstable blood flow and increase the risk of thrombosis. Therefore, in the calculation formula, MPCI is negatively correlated with the pressure gradient. A larger pressure gradient indicates a smaller MPCI value, indicating a potentially increased risk of thrombosis.

[0045] Furthermore, performing a logarithmic transformation on the cumulative heparin injection volume smoothed the data distribution, preventing large heparin doses from significantly influencing the formula results. The addition of +1 to the cumulative heparin injection volume ensured that the logarithmic value was also 0 even when the cumulative heparin injection volume was 0, maintaining calculation continuity. The MPCI was positively correlated with this logarithmic value, reflecting the positive effect of anticoagulant therapy in reducing thrombotic risk.

[0046] Calculation module 12 applies these three calculation results to the MPCI formula, achieving a comprehensive assessment of catheter pump efficiency and thrombosis risk. A high MPCI index indicates that the catheter pump is operating efficiently at high flow and high speed, and that anticoagulation therapy is adequate, resulting in a low thrombosis risk. Conversely, a low MPCI index indicates that catheter pump efficiency may be reduced, or anticoagulation therapy may be inadequate, leading to a correspondingly increased thrombosis risk.

[0047] Preferably, when processing continuously acquired data, the calculation module 12 calculates the corresponding MPCI value for each time point (preferably every 30 ms), forming an MPCI time series. This continuous calculation can reflect the dynamic changes in the mechanical-physiological coupling state during treatment, providing detailed temporal characteristics for subsequent thrombosis risk assessment. The calculation module 12 can perform statistical analysis on this MPCI time series data, calculating statistical quantities such as the MPCI mean within a specific time window, so that the subsequent thrombosis prediction module 13 can more comprehensively describe the coupling between the catheter pump operation and the patient's physiological state.

[0048] In a preferred embodiment, the calculation module 12 is further configured to: respectively determine the catheter pump flow fluctuation rate, the left ventricular pressure rising slope, the current rising slope and the mean of the aortic pressure; and determine a blood stasis index based on the correlation between the catheter pump flow fluctuation rate, the left ventricular pressure rising slope, the current rising slope and the mean of the aortic pressure; wherein the blood stasis index is positively correlated with the flow fluctuation rate, the ventricular pressure rising slope and the current rising slope, and is negatively correlated with the mean of the aortic pressure.

[0049] The blood stasis index (BSI) is designed to assess changes in pressure and speed during catheter pump operation, as well as blood flow stability, to determine the risk of blood stasis and thrombosis. The specific calculation formula is as follows:

[0050]

[0051] Specifically, the catheter pump flow fluctuation rate reflects the stability of blood flow. This parameter is the ratio of the standard deviation of the flow rate to the mean flow rate during a period of stable operation of the catheter pump at a certain gear. Its calculation formula is:

[0052]

[0053] The flow standard deviation σ is calculated by the following formula:

[0054]

[0055] The mean flow rate μ is calculated by the following formula:

[0056]

[0057] In the above formula, N represents the total number of sampling points in the selected time window, Q i represents the instantaneous flow value at the i-th sampling point. A higher flow fluctuation indicates more unstable blood flow, which increases the risk of blood stasis and thrombosis. Therefore, BSI is positively correlated with flow fluctuation.

[0058] The slope of left ventricular pressure rise reflects changes in cardiac contractile function and hemodynamics, and its calculation formula is:

[0059]

[0060] The calculation module 12 identifies the valley point (time point t1) and the peak point (time point t2) in the ventricular pressure waveform and calculates the left ventricular pressure P at time t2. Lv (t2) and left ventricular pressure P at t1 Lv The ratio of the pressure difference (t1) to the time difference yields the slope of the rise. A high slope may indicate cardiac overcompensation or increased blood shear stress, increasing the risk of thrombosis. Therefore, BSI is positively correlated with the slope of ventricular pressure rise.

[0061] Furthermore, the current rising slope reflects the change in the operating state of the catheter pump and the internal resistance. The calculation formula is:

[0062]

[0063] Preferably, the calculation module 12 selects two time points during a period of significant current I change (e.g., during the speed adjustment period): point t1, before the speed begins to change, and point t2, after the speed stabilizes. The calculation module 12 calculates the ratio of the current difference to the time difference between these two points. An abnormally high current slope may indicate increased resistance within the pump, such as thrombosis or mechanical failure. Therefore, the BSI is positively correlated with the current slope.

[0064] Furthermore, the mean aortic pressure reflects systemic blood perfusion. Preferably, the mean aortic pressure is calculated by taking the arithmetic average of aortic pressure data within a specific time window. A higher mean aortic pressure generally indicates good systemic blood perfusion and a lower risk of blood stasis, while a lower mean aortic pressure generally indicates a higher risk of blood stasis. Therefore, in the BSI calculation formula, BSI and mean aortic pressure are negatively correlated, and mean aortic pressure appears as the denominator in the calculation formula.

[0065] Calculation module 12 applies these four calculation results to the BSI formula, achieving a comprehensive assessment of blood stasis and thrombosis risk. Specifically, a high BSI index indicates unstable blood flow, cardiac overcompensation, or increased pump resistance, which may increase the risk of thrombosis. Conversely, a low BSI index indicates relatively stable blood flow and a relatively low risk of thrombosis.

[0066] Because catheter pumps may exhibit varying operating characteristics and blood flow patterns at different speed settings, a gear-specific calculation can more accurately assess thrombosis risk under various operating conditions. Therefore, in a preferred embodiment, calculation module 12 calculates BSI values ​​for each speed setting when processing data. Preferably, calculation module 12 also records BSI peak values ​​at each speed setting. These peak values ​​represent periods of peak blood retention and are closely associated with thrombosis trigger points, providing important early warning value in subsequent thrombosis risk assessments.

[0067] In a preferred embodiment, the thrombosis prediction module 13 is configured to: determine the mean value of the mechanical-physiological coupling index (MPCI), the peak value of the blood stasis index (BSI), and the cumulative injection volume of heparin respectively; input the mean value of the mechanical-physiological coupling index, the peak value of the blood stasis index, and the cumulative injection volume of heparin into the thrombosis prediction model to obtain the probability of thrombosis formation.

[0068] In a preferred embodiment, the thrombus prediction module 13 first acquires data from a preset time period before the end of the interventional ventricular assist device operation, preferably selecting data from one hour before the end of the procedure as the analysis sample. This data more accurately reflects the patient's immediate postoperative physiological state and thrombus risk. The thrombus prediction module 13 divides the acquired data into time windows and extracts features to ensure the reliability and clinical relevance of the prediction results.

[0069] In a preferred embodiment, the mechanical-physiological coupling index (MPCI) at each time point is determined according to a preset time interval, and the average value is taken as the mean mechanical-physiological coupling index. Preferably, the time interval is 30ms; the mean mechanical-physiological coupling index represents the average efficiency of the entire operation, reflects the overall coupling between the operation of the catheter pump and the patient's physiological state, and is an important indicator for assessing the risk of thrombosis.

[0070] In a preferred embodiment, the thrombus prediction module 13 is further configured to identify different speed gears in the data and divide the data into multiple time windows according to the gears. For each time window, the thrombus prediction module 13 calculates the corresponding blood stasis index (BSI) and then selects the maximum value from all calculated results as the peak value of the blood stasis index. Those skilled in the art will understand that the reason for selecting the peak value rather than the mean value is that the peak value of the blood stasis index represents the most severe period of blood retention, is closely related to the thrombus trigger point, and has a higher warning value. Therefore, even a short-term high BSI state may trigger thrombosis, and therefore the peak value is more suitable as a risk assessment indicator than the mean value.

[0071] In a preferred embodiment, the thrombosis prediction module 13 directly extracts the cumulative heparin injection volume from the data records. This value reflects the total amount of anticoagulant therapy administered during surgery and is another key indicator for assessing thrombosis risk. Specifically, because heparin is an anticoagulant, its cumulative dosage is negatively correlated with thrombosis risk. A higher dosage is associated with a lower thrombosis risk, but this balance also needs to be considered with respect to bleeding risk.

[0072] In a preferred embodiment, the thrombosis prediction model is configured as a logistic regression model. The model uses the mean value of the mechanical-physiological coupling index, the peak value of the blood stasis index, and the cumulative amount of heparin injected as input variables, and maps the result of the linear combination to the interval [0, 1] using a sigmoid function to represent the probability of thrombosis. Preferably, the mathematical expression of the logistic regression model is:

[0073]

[0074] The above formula contains four key parameters, namely the baseline risk b and three weight coefficients ω1, ω2, and ω3, among which the baseline risk b represents the initial risk when all input features are 0; the first weight coefficient ω1 is used to characterize the influence weight of the mean value of the mechanical-physiological coupling index, and takes a negative value, such as -0.8, indicating that the higher the mean value of the mechanical-physiological coupling index, the lower the thrombosis risk; the second weight coefficient ω2 is used to characterize the influence weight of the peak value of the blood stasis index, and takes a positive value, such as 1.5, indicating that the higher the peak value of the blood stasis index, the higher the thrombosis risk; the third weight coefficient ω3 is used to characterize the influence weight of the cumulative heparin injection amount, and takes a negative value, such as -0.002, indicating that the larger the heparin dosage, the lower the thrombosis risk.

[0075] Optionally, the three weight coefficients ω1, ω2, and ω3 can be obtained through historical clinical data, simulation model training, etc., and are not specifically limited here.

[0076] After calculating the probability of thrombosis, the thrombosis prediction module 13 presents the result as a numerical value between 0 and 1, making it easier for clinicians to intuitively understand and judge. For example, a predicted probability of 0.75 indicates a 75% chance of postoperative thrombosis, necessitating more aggressive preventive measures; a predicted probability of 0.15 indicates a lower risk of thrombosis, allowing conventional postoperative management. This quantitative risk assessment approach can help physicians develop personalized postoperative anticoagulation plans and improve patient outcomes.

[0077] In a preferred embodiment, the postoperative thrombosis prediction device further includes a verification module 14, which is designed to ensure the reliability of the prediction results. By comparing and analyzing the differences in parameters between the ventricular assist device operating in vivo and under in vitro control conditions, it identifies and eliminates interfering factors that may be caused by device abnormalities, thereby improving the accuracy of thrombosis prediction. Preferably, verification module 14 is configured to: obtain control operating parameters of the ventricular assist device in vitro in still water; detect the degree of match between the control operating parameters and the device operating parameters; and mark the corresponding data as invalid when the match is less than a preset threshold.

[0078] Specifically, after the catheter pump completes its treatment task and is removed from the patient's body, it is placed in a standard in vitro static water environment for a control test. During the static water test, the verification module 14 controls the catheter pump to operate in sequence according to the same gear speed used during the treatment process, and collects key operating parameters such as speed, current, and flow at each gear as control benchmark data. These control operating parameters reflect the standard working characteristics of the catheter pump in the absence of thrombosis and mechanical failure.

[0079] Furthermore, the verification module 14 calculates the matching degree between the control operating parameters and the equipment operating parameters actually collected during the treatment process. The matching degree calculation takes into account the differences in multiple key parameters, mainly including current differences, flow differences, changes in the pressure-flow relationship, etc. under the same speed conditions. The verification module 14 quantifies these differences into matching degree scores. The higher the score, the closer the actual operating state is to the ideal state, and the more normal the equipment working state is.

[0080] If the calculated match falls below a preset threshold, verification module 14 automatically marks the corresponding data as invalid, indicating a possible mechanical failure, sensor anomaly, or other non-physiological factor in the catheter pump. These anomalies can interfere with accurate assessment of thrombosis risk. For example, if the catheter pump's bearings are worn, even in the absence of thrombosis, they can cause an abnormally high current flow, falsely indicating a high thrombosis risk. Verification module 14's screening effectively eliminates these false positive results due to inherent device issues.

[0081] In a preferred embodiment, the verification module 14 is further configured to adjust or flag the output of the thrombus prediction module 13 based on the match analysis results. For data with a match well below a threshold, the verification module 14 not only marks the data as invalid but also generates a device abnormality warning, suggesting device inspection or replacement. For data with a match in the borderline range, the verification module 14 simply flags the data and reminds clinicians to consider the potential impact of device factors when interpreting the prediction results.

[0082] Through a systematic and automated verification process, the verification module 14 can objectively evaluate the impact of the device operating status on the prediction results, avoid misjudgment due to device factors, and improve the specificity of thrombosis prediction.

[0083] Specifically, the data acquisition module 11, calculation module 12, thrombus prediction module 13 and verification module 14 in the embodiment of the present invention are all integrated into the extracorporeal control device of the ventricular assist device. The extracorporeal control device serves as the core control unit of the entire ventricular assist system. It is not only responsible for the routine operation control of the catheter pump, but also simultaneously completes multiple tasks such as data acquisition, calculation analysis and prediction evaluation, and ultimately realizes the calculation of the mechanical-physiological coupling index and the blood stasis index, as well as the prediction of the probability of thrombosis.

[0084] An embodiment of the present invention further provides a ventricular assist system, which includes the extracorporeal control device in the above embodiment, a postoperative thrombus prediction device is integrated into the extracorporeal control device, and its corresponding functional modules are implemented through a software program.

[0085] refer to Figure 2 One embodiment of the present invention further provides a method for predicting postoperative thrombosis, comprising at least the following steps:

[0086] Step S21 , collecting device operating parameters, physiological parameters, and anticoagulation parameters during the operation of the interventional ventricular assist device.

[0087] In a preferred embodiment, the device operating parameters include the motor speed, current and catheter pump flow of the interventional ventricular assist device; the physiological parameters include left ventricular pressure and aortic pressure; and the anticoagulation parameters include the cumulative injection volume of heparin.

[0088] In a preferred embodiment, the method further includes a step of preprocessing the collected raw data, wherein the preprocessing includes at least data synchronization, data denoising and outlier detection. By preprocessing each parameter, the reliability of subsequent thrombosis risk assessment can be effectively improved.

[0089] Preferably, data collection uses a synchronization mechanism to ensure the time consistency of device operating parameters and physiological parameters. Preferably, the data collection frequency is set to once every 30ms, and each record contains a millisecond-accurate timestamp (YYYY-MM-DDHH:MM:SS.sss), as well as the device operating parameters and physiological parameters at the corresponding moment.

[0090] Furthermore, the synchronously collected parameters are organized into a structured columnar data format, where each row of data corresponds to a sampling time point, including the timestamp and all parameter values ​​of that time point, so that when a large amount of time series data is efficiently processed later, data in a specific time window can be quickly extracted for analysis.

[0091] Preferably, the left ventricular pressure and aortic pressure are processed using a second-order Butterworth low-pass filter with a cutoff frequency set to 5 Hz to remove high-frequency noise such as electromagnetic interference and catheter jitter to ensure data quality for subsequent analysis.

[0092] Step S22: Calculate a multidimensional evaluation index based on the device operating parameters, physiological parameters, and anticoagulation parameters.

[0093] Preferably, the multidimensional evaluation index includes a mechanical-physiological coupling index and a blood stasis index, wherein the mechanical-physiological coupling index is determined based on device operating parameters, physiological parameters and anticoagulation parameters, and the blood stasis index is determined based on device operating parameters and physiological parameters.

[0094] In a preferred embodiment, the method further includes step S221: respectively determining the product of the motor speed and the catheter pump flow rate, the pressure difference between the aortic pressure and the left ventricular pressure, and the logarithmic value of the cumulative heparin injection volume; and determining a mechanical-physiological coupling index based on the correlation between the product, the pressure difference, and the logarithmic value; wherein the mechanical-physiological coupling index is positively correlated with the product and the logarithmic value, and negatively correlated with the pressure difference.

[0095] In a preferred embodiment, the mechanical-physiological coupling index (MPCI) is calculated as follows:

[0096]

[0097] The product of the motor speed and the catheter pump flow rate reflects the mechanical output power and working status of the catheter pump. The pressure difference between the aortic pressure and the left ventricular pressure, namely the transvalvular pressure difference, reflects the resistance that the heart needs to overcome. Logarithmic transformation of the cumulative heparin injection volume can smooth the data distribution and avoid excessive influence of large doses of heparin on the formula results.

[0098] Finally, the three calculation results are combined in the MPCI formula to achieve a comprehensive assessment of catheter pump efficiency and thrombosis risk. A high MPCI index indicates that the catheter pump is operating efficiently at high flow and high speed, and that anticoagulation therapy is adequate, resulting in a low thrombosis risk. Conversely, a low MPCI index indicates that the catheter pump efficiency may be reduced, or anticoagulation therapy may be inadequate, leading to a correspondingly increased thrombosis risk.

[0099] In a preferred embodiment, step S222 is also included: determining the mean of the catheter pump flow fluctuation rate, the left ventricular pressure rising slope, the current rising slope and the aortic pressure respectively; determining the blood stasis index based on the correlation between the catheter pump flow fluctuation rate, the left ventricular pressure rising slope, the current rising slope and the aortic pressure mean; wherein the blood stasis index is positively correlated with the flow fluctuation rate, the ventricular pressure rising slope and the current rising slope, and is negatively correlated with the aortic pressure mean.

[0100] Preferably, the blood stasis index (BSI) is intended to assess the changes in pressure and speed during the operation of the catheter pump, as well as the stability of blood flow, so as to determine the risk of blood stasis and thrombosis. The specific calculation formula is as follows:

[0101]

[0102] Specifically, the catheter pump flow fluctuation rate reflects the stability of blood flow, and its calculation formula is:

[0103]

[0104] In the above formula, the flow standard deviation Traffic mean N represents the total number of sampling points in the selected time window, Q i Represents the instantaneous flow value of the i-th sampling point.

[0105] The slope of left ventricular pressure rise reflects changes in cardiac contractile function and hemodynamics, and its calculation formula is:

[0106]

[0107] By identifying the trough point (t1 time point) and the peak point (t2 time point) in the ventricular pressure waveform and calculating the ratio of the left ventricular pressure difference to the time difference between these two points, the rising slope can be obtained. A slope that is too high may indicate excessive cardiac compensation or increased blood flow shear force, increasing the risk of thrombosis.

[0108] Furthermore, the current rising slope reflects the change in the operating state of the catheter pump and the internal resistance. The calculation formula is:

[0109]

[0110] Specifically, the current rise slope is determined by selecting two time points during a period of significant current change (e.g., during speed adjustment): point t1, before the speed begins to change, and point t2, after the speed stabilizes. The ratio of the current difference to the time difference between these two points is then calculated. An abnormally high current rise slope may indicate increased resistance within the pump, such as thrombosis or mechanical failure.

[0111] Furthermore, the mean aortic pressure reflects the state of systemic blood perfusion. Preferably, the mean aortic pressure is obtained by taking the arithmetic average of the aortic pressure data within a certain time window. A higher mean aortic pressure usually indicates good systemic blood perfusion and a lower risk of blood stasis, whereas a lower mean aortic pressure indicates a higher risk of blood stasis.

[0112] The above four calculation results are combined and applied to the BSI formula to achieve a comprehensive assessment of blood stasis and thrombosis risk. Specifically, a high BSI index indicates unstable blood flow, cardiac overcompensation, or increased pump resistance, which increases the risk of thrombosis. Conversely, a low BSI index indicates relatively stable blood flow and a relatively low risk of thrombosis.

[0113] Step S23 , estimating the probability of thrombosis formation using a preset thrombosis prediction model based on the multidimensional evaluation index.

[0114] Preferably, step S23 further specifically includes: determining the mean value of the mechanical-physiological coupling index (MPCI), the peak value of the blood stasis index (BSI) and the cumulative injection volume of heparin respectively; inputting the mean value of the mechanical-physiological coupling index, the peak value of the blood stasis index and the cumulative injection volume of heparin into the thrombosis prediction model to obtain the probability of thrombosis formation.

[0115] In a preferred embodiment, data within a preset time period before the end of the operation of the interventional ventricular assist device is first obtained, preferably data 1 hour before the end of the operation is selected as the analysis sample, and then the mechanical-physiological coupling index (MPCI) of each time point is determined according to the preset time interval, and the average value is taken as the mean mechanical-physiological coupling index. Preferably, the time interval is 30ms.

[0116] In a preferred embodiment, step S23 further specifically includes identifying different speed gears in the data, and dividing the data into multiple time windows according to the gears. For each time window, the corresponding blood stasis index (BSI) is calculated, and then the maximum value is selected from all calculated results as the peak value of the blood stasis index.

[0117] In a preferred embodiment, the cumulative amount of heparin injected is directly extracted from the data record. This value reflects the total amount of anticoagulation therapy during the operation and is another key indicator for evaluating thrombosis risk.

[0118] In a preferred embodiment, the thrombosis prediction model is configured as a logistic regression model. The model uses the mean value of the mechanical-physiological coupling index, the peak value of the blood stasis index, and the cumulative amount of heparin injected as input variables, and maps the result of the linear combination to the interval [0, 1] using a sigmoid function to represent the probability of thrombosis. Preferably, the mathematical expression of the logistic regression model is:

[0119]

[0120] The above formula contains four key parameters, namely the baseline risk b and three weight coefficients ω1, ω2, and ω3, among which the baseline risk b represents the initial risk when all input features are 0; the first weight coefficient ω1 is used to characterize the influence weight of the mean value of the mechanical-physiological coupling index, and takes a negative value, such as -0.8, indicating that the higher the mean value of the mechanical-physiological coupling index, the lower the thrombosis risk; the second weight coefficient ω2 is used to characterize the influence weight of the peak value of the blood stasis index, and takes a positive value, such as 1.5, indicating that the higher the peak value of the blood stasis index, the higher the thrombosis risk; the third weight coefficient ω3 is used to characterize the influence weight of the cumulative heparin injection amount, and takes a negative value, such as -0.002, indicating that the larger the heparin dosage, the lower the thrombosis risk.

[0121] After calculating the probability of thrombosis, the result is presented as a numerical value between 0 and 1, making it easier for clinicians to intuitively understand and judge. For example, a predicted probability of 0.75 indicates that the patient has a 75% chance of developing a thrombosis after surgery, requiring more aggressive preventive measures. A predicted probability of 0.15 indicates that the patient's thrombosis risk is low, and conventional postoperative management plans can be adopted. This quantitative risk assessment method can help doctors develop personalized postoperative anticoagulation plans and improve patient prognosis.

[0122] In a preferred embodiment, the method further includes step S24: obtaining control operating parameters of the interventional ventricular assist device in extracorporeal static water; detecting the degree of matching between the control operating parameters and the device operating parameters; and marking the corresponding data as invalid data when the degree of matching is less than a preset threshold.

[0123] Specifically, after the catheter pump completes its treatment task and is removed from the patient's body, it is placed in a standard in vitro static hydrostatic environment for a control test. During the static hydrostatic test, the catheter pump is controlled to operate in sequence at the same gear speed used during the treatment process, and key operating parameters such as speed, current, and flow at each gear are collected as control benchmark data. These control operating parameters reflect the standard working characteristics of the catheter pump in the absence of thrombosis and mechanical failure.

[0124] Furthermore, by calculating the matching degree between the control operating parameters and the equipment operating parameters actually collected during the treatment process, the matching degree calculation takes into account the differences in multiple key parameters, mainly including current differences, flow differences, changes in the pressure-flow relationship, etc. under the same speed conditions. These differences are then quantified into matching degree scores. The higher the score, the closer the actual operating state is to the ideal state, and the more normal the equipment working state is.

[0125] When the calculated matching degree is less than the preset threshold, the corresponding data is marked as invalid data. This indicates that the catheter pump may have an abnormality caused by mechanical failure, sensor abnormality, or other non-physiological factors. These abnormalities will interfere with the accurate assessment of thrombosis risk. For example, if the bearings of the catheter pump are worn, even if there is no thrombosis, it may cause an abnormal increase in current, which may falsely indicate a high thrombosis risk.

[0126] Specifically, the present method embodiment and the aforementioned device embodiment originate from the same inventive concept, and many technical features are no longer described one by one, and can be naturally inherited in the present embodiment.

[0127] One embodiment of the present application provides a computer-readable storage medium, wherein the storage medium stores at least one instruction, and the at least one instruction is loaded and executed by a processor to implement the postoperative thrombosis prediction method as described above.

[0128] An embodiment of the present application provides an electronic device, comprising a processor and a memory, wherein the memory stores at least one instruction, and the instruction is loaded and executed by the processor to implement the postoperative thrombosis prediction method as described above.

[0129] Those skilled in the art will understand that all or part of the steps to implement the above embodiments may be accomplished by hardware, or may be accomplished by a program to instruct the relevant hardware, and the program may be stored in a computer-readable storage medium, which may be a read-only memory, a disk, or an optical disk, etc.

[0130] The embodiments of the present invention are described above in conjunction with the accompanying drawings, but the present invention is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the guidance of the present invention, ordinary technicians in this field can also make many forms without departing from the scope of protection of the present invention and the claims, all of which are protected by the present invention.

Claims

1. A postoperative thrombosis prediction device, characterized in that: include: a data acquisition module configured to acquire device operating parameters, physiological parameters, and anticoagulation parameters during operation of the interventional ventricular assist device; a calculation module configured to calculate a multidimensional assessment index based on the device operating parameters, the physiological parameters, and the anticoagulation parameters; The thrombus prediction module is configured to estimate the probability of thrombosis according to the multidimensional evaluation index through a preset thrombus prediction model.

2. The postoperative thrombosis prediction device according to claim 1, characterized in that: The device operating parameters include motor speed, current, and catheter pump flow of the interventional ventricular assist device; The physiological parameters include left ventricular pressure and aortic pressure; The anticoagulation parameters include the cumulative amount of heparin injected.

3. The postoperative thrombosis prediction device according to claim 2, characterized in that: The multidimensional assessment index includes a mechanical-physiological coupling index; The calculation module is further configured to determine the mechanical-physiological coupling index based on the device operating parameters, the physiological parameters, and the anticoagulation parameters.

4. The postoperative thrombosis prediction device according to claim 3, characterized in that: The calculation module is further configured to: respectively determining the product of the motor speed and the catheter pump flow rate, the pressure difference between the aortic pressure and the left ventricular pressure, and the logarithmic value of the cumulative heparin injection amount; The mechanical-physiological coupling index is determined based on the correlation between the product, the pressure difference and the logarithmic value; wherein the mechanical-physiological coupling index is positively correlated with the product and the logarithmic value, and negatively correlated with the pressure difference.

5. The postoperative thrombosis prediction device according to claim 4, characterized in that: The multidimensional assessment index also includes a blood stasis index; The calculation module is further configured to determine the blood stasis index based on the device operating parameters and the physiological parameters.

6. The postoperative thrombosis prediction device according to claim 5, characterized in that: The calculation module is configured as follows: The mean values ​​of the catheter pump flow fluctuation rate, left ventricular pressure rising slope, current rising slope and aortic pressure were determined respectively; The blood stasis index is determined based on the correlation between the catheter pump flow fluctuation rate, the left ventricular pressure rising slope, the current rising slope and the aortic pressure mean; wherein the blood stasis index is positively correlated with the flow fluctuation rate, the ventricular pressure rising slope and the current rising slope, and negatively correlated with the aortic pressure mean.

7. The postoperative thrombosis prediction device according to claim 6, characterized in that: The thrombus prediction module is configured as follows: respectively determining the mean value of the mechanical-physiological coupling index, the peak value of the blood stasis index, and the cumulative injection volume of heparin; The mean value of the mechanical-physiological coupling index, the peak value of the blood stasis index, and the cumulative injection amount of heparin are input into the thrombosis prediction model to obtain the probability of thrombosis formation.

8. The postoperative thrombosis prediction device according to claim 7, characterized in that: The thrombus prediction module is further configured to: Acquiring data within a preset time period before the end of operation of an interventional ventricular assist device; Determine the mechanical-physiological coupling index at each time point according to a preset time interval, and take the average value as the mechanical-physiological coupling index mean; The data at different speed gears are divided into multiple time windows, the blood stasis index of each time window is calculated respectively, and the maximum value is taken as the peak value of the blood stasis index.

9. The postoperative thrombosis prediction device according to claim 7, characterized in that: The thrombus prediction model is configured as a logistic regression model; The logistic regression model includes a first weight coefficient, a second weight coefficient, and a third weight coefficient; the first weight coefficient and the third weight coefficient are negative values, indicating that the risk of thrombosis decreases as they increase; the second weight coefficient is positive, indicating that the risk of thrombosis increases as it increases; Among them, the first weight coefficient is used to characterize the influence weight of the mean value of the mechanical-physiological coupling index, the second weight coefficient is used to characterize the influence weight of the peak value of the blood stasis index, and the third weight coefficient is used to characterize the influence weight of the cumulative injection amount of heparin.

10. The postoperative thrombosis prediction device according to claim 4, characterized in that: It also includes a verification module, wherein the verification module is configured to: To obtain control operating parameters of interventional ventricular assist devices in extracorporeal still water; detecting a degree of matching between the control operating parameters and the device operating parameters; When the matching degree is less than a preset threshold, the corresponding data is marked as invalid data.

11. A method for predicting postoperative thrombosis, characterized in that: include: Collect device operating parameters, physiological parameters and anticoagulation parameters during the operation of interventional ventricular assist devices; Calculating a multidimensional assessment index based on the device operating parameters, the physiological parameters, and the anticoagulation parameters; According to the multidimensional evaluation index, the probability of thrombosis is estimated by a preset thrombosis prediction model.

12. A ventricular assist system, characterized in that: Comprising the postoperative thrombosis prediction device as described in any one of claims 1-10.

13. A computer-readable storage medium, characterized in that The storage medium stores at least one instruction, which is loaded and executed by the processor to implement the postoperative thrombosis prediction method according to claim 11.

14. An electronic device, characterized in that: The electronic device includes a processor and a memory, wherein the memory stores at least one instruction, and the instruction is loaded and executed by the processor to implement the postoperative thrombosis prediction method according to claim 11.

Images