A dynamic cardiac tamponade prediction method, device, medium, and program product based on multi-modal continuous monitoring data

Abstract

The embodiment of the specification provides a dynamic cardiac tamponade prediction method, device, medium and program product based on multi-modal continuous monitoring data, and relates to the field of intelligent medical treatment. The method comprises the following steps: receiving pericardial effusion related physiological data including at least ultrasonic parameters, electrocardio parameters, body surface temperature parameters and clinical sign data of a sample; extracting a current effusion volume parameter in real time according to the ultrasonic parameters, and calculating an effusion growth rate per unit time; quantitatively extracting a heart rate value, a percentage of a decrease amplitude of a QRS complex amplitude relative to a baseline and an alternating change type of the QRS complex amplitude according to the electrocardio parameters; analyzing a temperature change trend according to the body surface temperature parameters; inputting the effusion growth rate, the quantitative electrocardio parameters, the temperature change trend and the clinical sign data into a preset multi-modal fusion scoring model to output a real-time comprehensive risk score; and generating a dynamic early warning signal of a corresponding level according to the real-time comprehensive risk score to realize a prior prediction of a risk of cardiac tamponade.

Description

Technical Field

[0001] This invention relates to the field of intelligent healthcare, and more specifically, to a method, device, medium, and program product for predicting dynamic cardiac tamponade based on multimodal continuous monitoring data. Background Technology

[0002] Cardiac tamponade is a life-threatening emergency, essentially caused by the rapid or excessive accumulation of fluid in the pericardial cavity, leading to a sharp increase in intracardiac pressure. This pressure then compresses the heart, restricts ventricular diastolic filling, and ultimately results in a sharp decrease in cardiac output and circulatory failure. Pericardial effusion of varying degrees may occur in some postoperative cardiac patients. Rapidly accumulating effusions (e.g., an increase of >150 mL within 24 hours postoperatively), even if the volume does not meet the traditional ultrasound definition of "large," can induce hemodynamic instability even during the compensatory phase. Therefore, early, continuous, and dynamic identification of cardiac tamponade risk, rather than relying solely on a single diagnosis after symptom onset, has become a core requirement of perioperative monitoring.

[0003] Currently, the widely used clinical tools for assessing cardiac tamponade are all single-time-point static scales, which have systemic limitations: (1) Inherent defects in assessment models: The European Society of Cardiology (ESC) 2015 Guidelines for the diagnosis and management of pericardial diseases uses low voltage and electrical alternans on electrocardiogram as auxiliary diagnostic indicators, but only as qualitative descriptions (such as "QRS amplitude in limb leads < 5 mm"), without quantitative inclusion in the scoring system, and relies on manual interpretation, which cannot achieve automated continuous monitoring; the Duanmu four-variable score (Emergency Cardiac Tamponade Risk Prediction) is a quantitative model, but it is only applicable to a single assessment in an emergency setting. Its variables (such as effusion diameter and right ventricular collapse) are all static anatomical parameters, and do not involve the effusion growth rate, a key dynamic indicator that determines the occurrence of tamponade (see European Heart Journal: Acute Cardiovascular Care 2021); the CTCAE classification and prognostic score for malignant effusion (LENT / PROMISE) focus on adverse event reporting or long-term prognosis, and are not related to real-time perioperative warning.

[0004] (2) Parameter dimension is seriously insufficient: Existing tables are generally limited to ultrasound and clinical parameters, lacking the ability to quantify and analyze cardiac electrical activity. Ang et al. (Ang KP, Nordin RB, Lee SCY, Lee CY, Lu HT. Diagnostic value of electrocardiogram in cardiac tamponade. Med J Malaysia. 2019 Feb;74(1):51-56. PMID: 30846663.) confirmed in a study of 42 patients with cardiac tamponade that low voltage of QRS complex and sensitivity to sinus tachycardia reached 67%–69%, while electrical alternation specificity was as high as 100%. However, these parameters were not structured into numerical features that could be input into the model in the ESC and Duanmu systems. At the same time, chest wall skin temperature change is a reliable peripheral marker of cardiac tamponade leading to low cardiac output → peripheral vasoconstriction → perfusion impairment (2025 ESICM guidelines on circulatory shock and hemodynamic monitoring). (2025)DOI: 10.1007 / s00134-025-08137-z), which was not included in any of the existing tables.

[0005] In summary, existing technologies have failed to address the fundamental technical challenge of how to achieve highly specific and low-delay risk grading and early warning based on continuous postoperative monitoring data and through dynamic quantification and collaborative analysis of multiple physiological parameters before the onset of cardiac tamponade symptoms. Summary of the Invention

[0006] This invention aims to at least solve one of the technical problems existing in the prior art. To this end, this invention provides a method, device, medium, and program product for dynamic cardiac tamponade prediction based on multimodal continuous monitoring data; the method of this invention continuously collects multimodal data (ultrasound, electrocardiogram, temperature, and clinical data), assesses the risk of cardiac tamponade in real time, and outputs a three-level warning (Level 1 Yellow Warning, Level 2 Orange Alert, Level 3 Red Emergency).

[0007] The first aspect of this application discloses a dynamic cardiac tamponade prediction method based on multimodal continuous monitoring data, comprising: Receive physiological data related to pericardial effusion from the collected samples, including at least ultrasound parameters, electrocardiogram parameters, body surface temperature parameters, and clinical signs data; Based on the ultrasound parameters, the current fluid volume parameters are extracted in real time, and the fluid growth rate per unit time is calculated. Based on the electrocardiogram parameters, the heart rate value, the percentage decrease in QRS complex amplitude relative to the baseline, and the alternation type of QRS complex amplitude are extracted quantitatively. The temperature change trend of chest wall skin temperature relative to baseline temperature was obtained based on body surface temperature parameters. Based on the effusion growth rate, quantified electrocardiogram parameters, temperature change trend, and clinical signs data, a preset multimodal fusion scoring model is input to output a real-time comprehensive risk score. Based on real-time comprehensive risk scores, dynamic early warning signals of corresponding levels are generated to enable advance prediction of the risk of cardiac tamponade.

[0008] The second aspect of this application discloses a dynamic cardiac tamponade prediction system based on multimodal continuous monitoring data, used to implement the method disclosed in the first aspect of this application. The system includes: The multi-source data acquisition module is used to acquire ultrasound parameters, electrocardiogram parameters, body surface temperature parameters, and clinical signs data; The dynamic parameter extraction module is used to perform calculations of effusion growth rate, quantification of electrocardiogram parameters, and analysis of changes in body surface temperature. The multimodal fusion scoring module incorporates the multimodal fusion scoring model disclosed in the first aspect of this application, used to generate a comprehensive risk score; The early warning module is used to output graded early warning signals based on the comprehensive risk score.

[0009] A third aspect of this application discloses a computer device, comprising: a memory and a processor; the memory is used to store a computer program; the processor executes the computer program to implement the steps of the above-described method.

[0010] The fourth aspect of this application discloses a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the above-described method.

[0011] The fifth aspect of this application discloses a computer program product, including a computer program that, when executed by a processor, implements the steps of the above-described method.

[0012] This application has the following beneficial effects: 1. This application addresses the limitations of existing scales that rely solely on static effusion volume (such as diameter and volume) by introducing the effusion growth rate (mL / h) as an independent scoring parameter in the ultrasound dimension. This parameter directly reflects the dynamic process of effusion accumulation, enabling the system to identify abnormal growth trends of "above 10 mL / h" even when the patient has no clinical symptoms such as chest tightness or hypotension (i.e., the compensatory period). This advances the warning time window to several hours before cardiac tamponade occurs, overcoming the shortcomings of existing technologies that tend to rely on "post-hoc diagnosis and passive response."

[0013] 2. This application incorporates the percentage decrease in QRS amplitude (relative to individual baseline) and electrical alternation characteristic markers (binary Boolean value, present or absent) in the electrocardiogram signal as structured inputs into a multimodal model. This design utilizes the high sensitivity (67%–69%) of low QRS voltage confirmed by the Ang study for primary screening, and also uses the 100% specificity of electrical alternation to strongly confirm warnings of level two and above, significantly improving the positive predictive value (PPV) and clinical reliability of multi-parameter joint discrimination, and avoiding false alarms caused by fluctuations in a single parameter.

[0014] 3. This application incorporates the trend of chest wall skin temperature change as an auxiliary weighting dimension into the scoring model, which has a solid physiological basis (cardiac tamponade → low cardiac output → sympathetic excitation → peripheral vasoconstriction → decreased skin temperature), and forms a "central-peripheral" dual-axis verification logic with ultrasound and electrocardiogram parameters. This approach enhances the system's ability to capture early compensatory physiological disturbances, especially in the period when ultrasound findings are not yet typical (e.g., right ventricular collapse is not obvious) but peripheral perfusion impairment has already occurred, providing crucial supplementary evidence.

[0015] 4. This application generates a continuous comprehensive risk score by inputting four dimensions of data into the model: ultrasound parameters (fluid volume + rate + signs of tamponade), electrocardiogram parameters (heart rate + QRS decrease % + electrical alternans), body surface temperature parameters (skin temperature change trend), and clinical data (blood pressure, SpO2, NRS score). Based on the score, it outputs a three-level dynamic early warning signal, realizing the update from manual single-point interpretation to automatic multi-dimensional collaborative decision-making by machines. Furthermore, the system disclosed in this application can be seamlessly embedded into the postoperative monitoring process, supporting preoperative pre-setting and continuous operation for 72 hours after surgery, achieving the goal of "early warning, graded intervention, and precise decision-making". Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 This is a schematic diagram of the method flow provided in the first aspect of the present invention; Figure 2 This is a schematic diagram of a dynamic cardiac tamponade prediction system based on multimodal continuous monitoring data provided in the second aspect of the present invention. Detailed Implementation

[0018] To enable those skilled in the art to better understand the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings.

[0019] In some of the processes described in the specification, claims, and accompanying drawings of this invention, multiple operations appearing in a specific order are included. However, it should be clearly understood that these operations may not be executed in the order they appear herein, or may be executed in parallel. The operation numbers, such as 101, 102, etc., are merely used to distinguish different operations and do not represent any execution order. Furthermore, these processes may include more or fewer operations, and these operations may be executed sequentially or in parallel. It should be noted that the descriptions such as "first," "second," etc., in this document are used to distinguish different messages, devices, modules, etc., and do not represent a sequential order, nor do they limit "first" and "second" to different types.

[0020] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0021] Figure 1 This is a schematic flowchart of a dynamic cardiac tamponade prediction method based on multimodal continuous monitoring data provided by an embodiment of the present invention. Specifically, the method includes the following steps: S101: Receive physiological data related to pericardial effusion from the collected sample, wherein the physiological data includes at least ultrasound parameters, electrocardiogram parameters, body surface temperature parameters, and clinical signs data; In some embodiments, the terms “subject” or “test subject” or “sample” as used herein refer to any animal (e.g., a mammal), including but not limited to humans, non-human primates, rodents, etc., which will become the recipient of a particular treatment. Generally, the terms “subject” and “patient” are used interchangeably herein when referring to human subjects. Preferably, the sample is a human.

[0022] In some embodiments, the clinical signs data include any one or more of the following: systolic blood pressure, oxygen saturation (SpO2 < 92%), respiratory rate, NRS pain score (chest tightness / chest pain), level of consciousness assessment, and orthopnea performance.

[0023] In some specific embodiments, systolic blood pressure <100 mmHg is a validated independent predictor in the Duanmu score (assigned a value of 1.5 points), and this scale uses a lower threshold (<90 mmHg) as the trigger condition for a Level 3 emergency.

[0024] S102: Extract the current fluid volume parameters in real time based on the ultrasound parameters, and calculate the fluid growth rate per unit time; In some embodiments, the effusion growth rate is an instantaneous rate parameter obtained by differential calculation of the estimated width of the diastolic anechoic zone or the effusion volume over a continuous time period using the sliding time window method. This parameter is a dynamic parameter not disclosed in existing cardiac tamponade assessment scales.

[0025] In some specific embodiments, real-time ultrasound measurement is used to quantify the fluid volume to obtain the fluid amount (ml or diastolic anechoic zone width mm). The fluid growth rate is measured in ml / h, and it is obtained by continuously monitoring and recalculating the change in fluid volume per unit time. This parameter is an innovative dimension that is completely missing from existing tables.

[0026] In some embodiments, the ultrasound parameters also include: ultrasound signs of tamponade, such as right atrial systolic collapse, right ventricular diastolic collapse, and inferior vena cava dilation without respiratory variability, which are typical echocardiographic manifestations of cardiac tamponade and are used as escalation criteria for grading decisions.

[0027] S103: Quantitatively extract the heart rate value, the percentage decrease in QRS complex amplitude relative to the baseline, and the alternation type of QRS complex amplitude based on the electrocardiogram parameters; In some embodiments, the percentage decrease in QRS complex amplitude relative to baseline is determined by calculating the baseline QRS amplitude value established during the stable period of the sample and the real-time monitoring value. In some more specific embodiments, the electrocardiogram parameters are processed as follows: Sinus tachycardia: Heart rate > 100 bpm, quantified heart rate value, one of the triggering conditions for a Level 1 warning.

[0028] Low voltage (QRS amplitude reduction percentage): Referring to the diagnostic criteria for low voltage in the study by Ang et al. (QRS < 5 mm in limb leads or QRS < 10 mm in precordial leads), and combined with baseline measurement data, the reduction percentage was calculated (>30% for Level II alert, >50% for Level III emergency). In this embodiment, the QRS amplitude reduction percentage is the relative amplitude reduction value, calculated as follows: Relative amplitude reduction value (percentage reduction) = (maximum amplitude - minimum amplitude) ÷ maximum amplitude × 100%.

[0029] Amplitude usually refers to the voltage (μV) value of a QRS complex (or a specific waveform, such as the amplitude of an HFQRS wave). The maximum and minimum values ​​are taken from the amplitude extremes of a specific time interval (such as the first 3 minutes of the exercise period).

[0030] The QRS complex is an important waveform in an electrocardiogram (ECG) that reflects ventricular depolarization activity. The QRS complex is a combination of three closely linked electrical potential waveforms, representing: Q wave: The first downward wave in the QRS complex (before the R wave).

[0031] R wave: The first upward wave in the QRS complex (it is the highest wave in the complex and represents the main part of ventricular depolarization).

[0032] S wave: The first downward wave after the R wave.

[0033] In some embodiments, the QRS amplitude alternation type is a classification based on the presence or absence of QRS amplitude alternation pattern recognition. Electrical alternation refers to alternating QRS amplitudes, a relatively specific sign of significant pericardial effusion. Ang's study showed a specificity of 100% but a sensitivity of only 5%, thus making it suitable as an auxiliary condition for escalating alerts rather than an independent indicator. (QRS electrical alternation refers to the beat-by-beat alternation of the morphology or amplitude of the QRS complex on an electrocardiogram, the most common form being the alternation of high-amplitude and low-amplitude heartbeats. It is unrelated to the respiratory cycle, persists after breath-holding, and often indicates severe organic heart disease or pericardial effusion.)

[0034] Compared to the qualitative descriptions in the ESC guidelines ("low voltage", "electrical alternation"), this system quantifies ECG parameters (heart rate bpm, QRS amplitude reduction percentage, presence or absence of electrical alternation) and incorporates them into an automated multimodal scoring system, achieving a functional upgrade of ECG parameters from "qualitative auxiliary diagnosis" to "quantitative scoring dimension".

[0035] As an example, the thresholds for the rate of fluid accumulation are described; rapid accumulation (150–250 mL) is a typical threshold for the occurrence of acute cardiac tamponade. The thresholds for Level I warning (>10 ml / h), Level II alert (>20 ml / h), and Level III emergency (>30 ml / h) in this application are based on the following clinical reasoning: 1. Level III Emergency (>30 ml / h): If this rate of accumulation continues for 1 hour, the fluid volume can reach over 30 ml. Although it has not yet reached the 150 ml tamponade threshold, in a continuous monitoring scenario, a sustained rate of 30 ml / h means that 45–60 ml can accumulate within 1.5–2 hours, and 120–150 ml within 4–5 hours, which is close to the clinical danger threshold. This threshold setting reflects a conservative safety margin—that is, a red emergency alert is issued before reaching the clinical tamponade threshold (while still in the compensatory phase).

[0036] 2. Level 2 Alert (>20 ml / h): Reaching 20 ml / h for 30 minutes indicates a significantly accelerated rate of fluid accumulation, requiring emergency assessment within 1 hour.

[0037] 3. Level 1 warning (>10 ml / h): This indicates a trend of fluid accumulation and requires clinical evaluation within 24 hours.

[0038] It is important to emphasize that the multi-dimensional parameters included in this application can be scaled for specific clinical applications. The effusion growth rate parameter of this scale is a continuous rate parameter under continuous monitoring scenarios, which is a dynamic parameter that existing scales such as the ESC score and Duanmu score do not possess. The Duanmu score uses the effusion diameter as a static assignment (0 points for 1–2 cm, 1.5 points for 2–3 cm, and 2 points for >3 cm), but has never introduced a rate dimension. Therefore, the innovative composition of the effusion growth rate parameter is a core function completely missing in existing scales, and it is also the technical foundation for this scale to achieve early warning.

[0039] S104: The temperature change trend of chest wall skin temperature relative to baseline temperature is obtained based on body surface temperature parameters; In some embodiments, the temperature change trend is continuously monitored by a contact or non-contact temperature sensor and participates in the multimodal fusion scoring in the form of temperature difference or change slope, which serves as an auxiliary weighted correction for cardiac tamponade risk under low perfusion conditions.

[0040] In some specific embodiments, the use of skin temperature as an auxiliary multimodal parameter is based on the following physiological basis: When cardiac tamponade occurs, the intracardiac pressure rises sharply, ventricular diastolic filling is restricted, cardiac output is drastically reduced, and the patient enters a state of hypotension and low cardiac output. This state triggers compensatory sympathetic responses, including increased heart rate and peripheral vasoconstriction (especially in the skin). Regarding peripheral perfusion assessment, in 2014, European experts recommended incorporating abnormal skin perfusion parameters into the definition and treatment of shock. Studies have shown that clinical assessment of skin temperature can identify potential severe organ dysfunction in hemodynamically stable patients. Therefore, decreased skin temperature is one of the peripheral manifestations of hypoperfusion caused by cardiac tamponade.

[0041] The positioning of skin temperature as an indicator in this application: Skin temperature is used as an auxiliary rather than a decisive indicator in this application. Its individual changes do not trigger puncture decisions. Only when it is abnormal in conjunction with major parameters such as ultrasound and electrocardiogram does it play a synergistic weighting role in multimodal fusion scoring, thereby enhancing the sensitivity of early warning. This application classifies skin temperature changes into: a decrease of <0.5℃ (normal / Level 1), a decrease of 0.5–1.5℃ (Level 2 alarm auxiliary prompt), and a decrease of >1.5℃ or obvious coldness of the extremities (Level 3 emergency auxiliary prompt).

[0042] S105: Based on the effusion growth rate, quantified electrocardiogram parameters, temperature change trend, and clinical signs data, a preset multimodal fusion scoring model is input to output a real-time comprehensive risk score; In some embodiments, the multimodal fusion scoring model is a statistical model or machine learning model trained through supervised learning. The inputs include the rate of effusion growth, quantified electrocardiographic parameters, temperature change trends, and clinical signs. The output is a continuous comprehensive risk score that reflects the early risk of cardiac tamponade. In some specific embodiments, the continuous score is between 0 and 1.

[0043] S106: Generate dynamic early warning signals of corresponding levels based on real-time comprehensive risk scores to achieve advance prediction of the risk of cardiac tamponade.

[0044] In some embodiments, the hierarchical definitions of dynamic early warning signals are shown in Table 1. Table 1. Definition of Hierarchy

[0045] In some embodiments, when using the multi-dimensional parameters of this application for early warning only, as an example, some quantitative thresholds are shown in Table 2. Table 2 Quantitative thresholds for multimodal parameters

[0046] Note: When any one of the criteria reaches Level 3, or two criteria reach Level 2 and are accompanied by clinical symptoms, the system will directly escalate to a red alert. Skin temperature, as an auxiliary parameter, does not trigger any level on its own, but plays a synergistic weighting role in multimodal fusion. Additionally, besides ultrasound directly indicating a Level 3 warning, other situations indicating a Level 3 warning may trigger enhanced ultrasound frequency, etc., to use ultrasound to more precisely verify whether there is a rapid increase in pericardial effusion.

[0047] In some embodiments, the pre-prediction results output based on the dynamic early warning signal include, but are not limited to, paper or electronic reports. These results are obtained by the intelligent machine based on the subject's relevant data and are intended only as a reference for medical personnel, not as the subject's final diagnostic result. The priority and alert level for ultrasound detection of fluid accumulation should be increased. For example, if ultrasound indicates a growth rate ≥10ml / h, even if it falls under the first-level warning category, a strong alert should still be issued to remind medical personnel to pay attention to the patient's condition and provide further information to the family.

[0048] In some embodiments, the thresholds involved in the scheme are obtained through training on training set samples. They can be specific thresholds or ranges, and the specific form is not specifically limited in this embodiment.

[0049] In some embodiments, a second aspect of this application discloses a dynamic cardiac tamponade prediction system based on multimodal continuous monitoring data, used to implement the method disclosed in the first aspect of this application, the system comprising: The multi-source data acquisition module is used to acquire ultrasound parameters, electrocardiogram parameters, body surface temperature parameters, and clinical signs data; The dynamic parameter extraction module is used to perform calculations of effusion growth rate, quantification of electrocardiogram parameters, and analysis of changes in body surface temperature. The multimodal fusion scoring module incorporates the multimodal fusion scoring model disclosed in the first aspect of this application, used to generate a comprehensive risk score; The early warning module is used to output graded early warning signals based on the comprehensive risk score.

[0050] In some embodiments, when the multi-dimensional parameters contained in this application are made into a scale for specific clinical applications, a systematic comparison between this scale and existing scales is as follows:

[0051] In some embodiments, the system also includes a module for automatically generating puncture suggestions upon triggering a red emergency alert. The following is the basis for the puncture path selection and related information in the module: 1. Puncture path selection: Ultrasound-guided pericardiocentesis is currently considered the standard treatment. In recent years, various puncture routes have been reported, mainly differing in the puncture site (subxiphoid, apex of the heart, or parasternal).

[0052] Apical approach: The puncture point is located at the fifth or sixth intercostal space along the left sternal border, 1–2 cm medial to the cardiac dullness border. The needle is inserted along the upper edge of the rib, pointing towards the right shoulder or spine. According to data from a 2022 study published in *The Ultrasound Journal*, the average thickness of effusion via the apical approach is 29 mm, with a depth of approximately 22 mm from the pericardial cavity. This approach has a significant advantage in terms of effusion thickness and is suitable for apical pericardial effusion.

[0053] Subxiphoid approach: The needle is inserted at the angle between the xiphoid process and the left costal arch, pointing towards the left shoulder at a 30°-45° angle to the abdominal wall, to a depth of approximately 64 mm. This approach is commonly used for pericardiocentesis in emergency unstable patients.

[0054] Parasternal approach: The effusion thickness along the left parasternal approach was approximately 10 mm, and the depth from the pericardial cavity was approximately 23 mm. Osman et al. reported a new technique for pericardiocentesis in the medial-lateral parasternal plane, which was successfully performed on 11 patients, with an average operation time of 309 ± 76.4 seconds and no operation-related complications.

[0055] 2. Safety Tips: Both the ESC 2015 guidelines and the MSD treatment manual state that ST-segment elevation should be monitored on an electrocardiogram during pericardiocentesis (indicating contact with the epicardium, requiring needle withdrawal). Real-time ultrasound guidance is currently considered the preferred image-guided method for pericardiocentesis.

[0056] 3. Post-traffic acquisition management: The MSD treatment manual states that if continuous drainage is required, a plastic catheter can be inserted through a puncture needle, and the catheter should be removed when the drainage volume is less than 25–50 ml / 24 hours (usually 2–4 days).

[0057] 4. Design of the puncture suggestion module in this application: Based on the aforementioned Chinese and foreign literature, this scale generates puncture recommendations when a red emergency is triggered, including: a recommended route (apical or subxiphoid approach) based on real-time ultrasound localization of the largest fluid pool in the effusion, estimated needle depth, safety tips (real-time ultrasound guidance, ECG ST-segment monitoring, cardiac surgery preparation), a pre-procedure preparation checklist, and post-drainage management recommendations. This module upgrades the scale from an "assessment tool" to a "clinical decision support system."

[0058] A third aspect of this application discloses a computer device, which may include: one or more processors and one or more memories; wherein the memories store computer-readable code that, when run by the one or more processors, can perform the methods described above.

[0059] The processor in this embodiment can be an integrated circuit chip with signal processing capabilities. The processor can be a general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), an off-the-shelf programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. It can implement or execute the methods, operations, and logic block diagrams disclosed in this embodiment. The general-purpose processor can be a microprocessor or any conventional processor, and can be based on an x86 or ARM architecture.

[0060] In general, the various exemplary embodiments of this disclosure can be implemented in hardware or dedicated circuitry, software, firmware, logic, or any combination thereof. Some aspects can be implemented in hardware, while others can be implemented in firmware or software that can be executed by a controller, microprocessor, or other computing device. When aspects of embodiments of this disclosure are illustrated or described as block diagrams, flowcharts, or using some other graphical representation, it will be understood that the blocks, apparatuses, systems, techniques, or methods described herein can be implemented as non-limiting examples in hardware, software, firmware, dedicated circuitry or logic, general-purpose hardware or controllers or other computing devices, or some combination thereof.

[0061] For example, the methods or apparatus according to embodiments of this disclosure can also be implemented using the architecture of a computing device. A computing device may include a bus, one or more CPUs, read-only memory (ROM), random access memory (RAM), a communication port connected to a network, input / output components, a hard disk, etc. Storage devices in the computing device, such as ROM or hard disk, may store various data or files used for processing and / or communication of the methods provided in this disclosure, as well as program instructions executed by the CPU. The computing device may also include a user interface. Of course, the architecture described above is merely exemplary, and one or more components of the computing device described above may be omitted as needed when implementing different devices.

[0062] This invention also includes a computer-readable storage medium storing computer-readable instructions. When the computer-readable instructions are executed by a processor, the methods disclosed in this embodiment can be performed. The computer-readable storage medium in this embodiment can be volatile memory or non-volatile memory, or may include both volatile and non-volatile memory. Non-volatile memory can be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. Volatile memory can be random access memory (RAM), which is used as an external cache. By way of example, but not limitation, many forms of RAM are available, such as static random access memory (SRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (SDRAM), double data rate synchronous dynamic random access memory (DDRSDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchronous interconnected dynamic random access memory (SLDRAM), and direct memory bus random access memory (DRRAM). It should be noted that the memory of the methods described herein is intended to include, but is not limited to, these and any other suitable types of memory.

[0063] This disclosure also provides a computer program product or system, including a computer program that, when executed by a processor, implements the steps of the above-described method. The computer program product or computer program is used to execute a dynamic cardiac tamponade prediction system based on multimodal continuous monitoring data. The system includes: The multi-source data acquisition module 201 is used to acquire ultrasound parameters, electrocardiogram parameters, body surface temperature parameters, and clinical signs data. The dynamic parameter extraction module 202 is used to perform calculation of effusion growth rate, quantification of electrocardiogram parameters, and analysis of body surface temperature change trends. The multimodal fusion scoring module 203 incorporates the multimodal fusion scoring model disclosed in the first aspect of this application, used to generate a comprehensive risk score; The early warning module 204 is used to output a graded early warning signal based on the comprehensive risk score.

[0064] The exemplary embodiments of this disclosure described in detail above are merely illustrative and not restrictive. Those skilled in the art will understand that various modifications and combinations can be made to these embodiments or their features without departing from the principles and spirit of this disclosure, and such modifications should fall within the scope of this disclosure.

Claims

1. A dynamic cardiac tamponade prediction method based on multimodal continuous monitoring data, characterized in that, include: Receive physiological data related to pericardial effusion from the collected samples, including at least ultrasound parameters, electrocardiogram parameters, body surface temperature parameters, and clinical signs data; Based on the ultrasound parameters, the current fluid volume parameters are extracted in real time, and the fluid growth rate per unit time is calculated. Based on the electrocardiogram parameters, the heart rate value, the percentage decrease in QRS complex amplitude relative to the baseline, and the alternation type of QRS complex amplitude are extracted quantitatively. The temperature change trend of chest wall skin temperature relative to baseline temperature was obtained based on body surface temperature parameters. Based on the effusion growth rate, quantified electrocardiogram parameters, temperature change trend, and clinical signs data, a preset multimodal fusion scoring model is input to output a real-time comprehensive risk score. Based on real-time comprehensive risk scores, dynamic early warning signals of corresponding levels are generated to enable advance prediction of the risk of cardiac tamponade.

2. The dynamic cardiac tamponade prediction method based on multimodal continuous monitoring data according to claim 1, characterized in that, The fluid growth rate is an instantaneous rate parameter obtained by differential calculation of the estimated width of the anechoic zone or the fluid volume during continuous diastole using the sliding time window method.

3. The dynamic cardiac tamponade prediction method based on multimodal continuous monitoring data according to claim 1, characterized in that, The percentage decrease in QRS complex amplitude relative to the baseline was determined by calculating the baseline QRS complex amplitude value established during the stable period of the sample and the real-time monitoring value. The alternating amplitude type of the QRS group is a classification of presence or absence based on the alternating amplitude pattern recognition of the QRS group.

4. The dynamic cardiac tamponade prediction method based on multimodal continuous monitoring data according to claim 1, characterized in that, Temperature change trends are continuously monitored by contact or non-contact temperature sensors and are used in multimodal fusion scoring in the form of temperature difference or change slope. Their role is to provide auxiliary weighted correction for cardiac tamponade risk under low perfusion conditions.

5. The dynamic cardiac tamponade prediction method based on multimodal continuous monitoring data according to claim 1, characterized in that, The clinical signs and symptoms data include any one or more of the following: systolic blood pressure, blood oxygen saturation, respiratory rate, NRS pain score, assessment of level of consciousness, and orthopnea.

6. The dynamic cardiac tamponade prediction method based on multimodal continuous monitoring data according to claim 1, characterized in that, The multimodal fusion scoring model is a statistical model or machine learning model trained through supervised learning. The inputs include the effusion growth rate, quantified electrocardiogram parameters, temperature change trends, and clinical signs data. The output is a continuous comprehensive risk score that can reflect the early risk of cardiac tamponade.

7. A dynamic cardiac tamponade prediction system based on multimodal continuous monitoring data, characterized in that, The system for implementing the method according to any one of claims 1-6 comprises: The multi-source data acquisition module is used to acquire ultrasound parameters, electrocardiogram parameters, body surface temperature parameters, and clinical signs data; The dynamic parameter extraction module is used to perform calculations of effusion growth rate, quantification of electrocardiogram parameters, and analysis of changes in body surface temperature. A multimodal fusion scoring module is used to generate a comprehensive risk score using the multimodal fusion scoring model according to claim 6; The early warning module is used to output graded early warning signals based on the comprehensive risk score.

8. A computer device, characterized in that, The device includes: a memory and a processor; the memory is used to store a computer program; the processor executes the computer program to implement the steps of the method according to any one of claims 1-6.

9. A computer-readable storage medium, characterized in that, It stores a computer program that, when executed by a processor, implements the steps of the method as described in any one of claims 1-6.

10. A computer program product, comprising a computer program, characterized in that, When executed by a processor, the computer program implements the steps of the method described in any one of claims 1-6.

Images