Invasive blood pressure monitor dynamic waveform blood pressure calibration system

Abstract

The application discloses a kind of invasive blood pressure monitor dynamic waveform blood pressure calibration system, comprising: pressure filing module, for obtaining pump stroke, valve opening and isolation chamber output pressure, generate pressure file calibration table;Fragment identification module, for collecting systolic peak, double beat incision and steep rising edge sampling section, generate resonance candidate fragment set. Two cooperation realizes dynamic waveform segmentation calibration. Fingerprint registration module is used to analyze the frequency of conduit resonance, oscillation start and stop pressure and peak amplitude decreasing state, and generate subfile resonance fingerprint table;Waveform checking module is used to compare high-frequency actuation output and standard waveform, and generate standard input waveform identification. Two realize deviation positioning under different pressure intervals. Compensation ware module is used to calculate the intensity difference between the detected output and standard input and the phase time difference, and establish dynamic calibration compensation table covering amplitude, frequency response and timing error.

Description

Technical Field

[0001] This invention relates to the field of signal pattern recognition technology, and in particular to a dynamic waveform blood pressure calibration system for an invasive blood pressure monitor. Background Technology

[0002] The field of signal pattern recognition technology mainly involves the acquisition, segmentation, feature extraction, waveform template comparison, peak and trough determination, and signal category identification of continuously changing pressure signals, voltage signals, pulse wave signals, and interference signals. This includes signal source determination, sampling frequency setting, waveform period division, amplitude change extraction, frequency component differentiation, abnormal band identification, and feature rule matching. The original signal is acquired using sensors, and recognition rules are established based on waveform amplitude, period, rising edge, falling edge, baseline drift, noise amplitude, and repeatability characteristics. The signal to be tested is then compared with preset signal samples for identification. Among them, the traditional dynamic waveform blood pressure calibration of invasive blood pressure monitors refers to the calibration of the dynamic pressure changes generated by the invasive blood pressure waveform directly measured by the arterial catheter under the influence of heartbeat, respiration, motion artifacts and vascular pulsation. It is used to calibrate systolic pressure, diastolic pressure, mean arterial pressure, pulse rate, pulse wave rising edge, pulse wave falling edge, pressure baseline drift, respiratory modulation waveform and motion artifact disturbance amplitude. The traditional method uses a standard pressure source to apply a fixed pressure, a pressure sensor to read the pressure value, a signal generator to output a corresponding voltage signal, and a waveform generator to output a fixed frequency pulsating pressure waveform. The calibration is based on the pressure amplitude, waveform period, pulse rate set value and voltage-pressure conversion relationship.

[0003] Currently, several blood pressure simulation and calibration devices are available on the market. Among them, Chinese patent CN103815885A discloses an invasive blood pressure testing device. This device mainly consists of a volume adjustable loop system, a sensor system, a dynamic pressure regulation system, and a control system. It relies on mechanical and hydraulic structures to simulate the static and dynamic blood pressure hydraulic waveforms of the human body. It mainly serves as a standard pressure source to complete the basic pressure test of the device under test. This device only realizes the generation and output of standard waveforms and does not have the functions of dynamic waveform error analysis, feature recognition, and deviation compensation.

[0004] Another Chinese patent, CN109805914A, discloses a dynamic calibration instrument for non-invasive blood pressure monitors. This device, with its core components including a gas path structure, a stepper motor, a piston, and supporting hardware drive circuitry, is designed for pressure and pulse simulation calibration of cuff-type non-invasive blood pressure devices. However, because non-invasive blood pressure monitors use gas as the transmission medium, their gas path pressure transmission and fluctuation characteristics are fundamentally different from those of invasive blood pressure monitors, which use liquid media and catheter fluid paths. Therefore, the hardware architecture and control logic of this device are not suitable for calibration scenarios involving dynamic waveforms of invasive blood pressure.

[0005] Existing technologies, in practical operation, largely rely on fixed-value pressure sources, fixed-frequency pulsating pressure waveforms, voltage-pressure conversion relationships, and single-point readings for calibration. The relationship between pressure application and the response of the monitored instrument tends to be static, making it difficult to reflect the transient coupling changes in the arterial catheter, sensor fluid circuit, isolation lumen elasticity, and connecting tubing under different pressure levels. Specifically, there are three independent drawbacks: 1. Static calibration cannot reflect dynamic waveform errors. Traditional methods apply static pressure using a fixed pressure source, relying on voltage-pressure conversion and single-point readings for calibration. This fails to simulate the dynamic waveform characteristics under conditions of rising heart rate, dicrotic notch, and respiratory modulation, leading to inconsistent calibration results in the high systolic blood pressure range and low diastolic blood pressure range. In the rising heart rate range, near the dicrotic notch, or in respiratory modulation scenarios, local high-frequency oscillations are often treated as general noise or artifacts, failing to reflect the transient coupling changes in the arterial catheter, sensor fluid path, and isolation chamber at different pressure levels, thus affecting the reliability of systolic blood pressure judgment and mean pressure calculation.

[0006] 2. Resonance characteristics are not tracked and located in stages. Existing systems treat high-frequency oscillations in dynamic waveforms as general noise or artifacts, lacking a staged tracking mechanism for the start and end positions, duration, peak decay trend, and pressure landing point of the oscillations. They do not record the start and end positions, duration, and peak decay trend of the oscillations according to pressure ranges, making it impossible to pinpoint whether the deviation originates from the high-pressure or low-pressure segment. This leads to waveform distortion, notch distortion, and inconsistent dynamic response in different pressure ranges, making it difficult to accurately locate the source of the deviation.

[0007] 3. Intensity and phase errors are not jointly compensated. Existing compensation methods only correct for a single dimension, either amplitude or frequency. Traditional fixed-frequency waveform output methods typically simplify catheter resonance to a periodic input response, ignoring the hysteresis relationship between the standard input intensity and the tested output intensity, as well as the coupling effect of phase shift. When rapid fluid resuscitation, changes in body position, or catheter bending occur, although the monitor can still maintain a pressure output close to the set value, waveform peak overshoot, lag in fallback, and local oscillation distortion still exist. This leads to waveform peak overshoot, lag in fallback, and local oscillation distortion, affecting the reliability of systolic blood pressure judgment, average pressure calculation, and alarm threshold. Summary of the Invention

[0008] To address the technical problems existing in the prior art, this invention provides a dynamic waveform blood pressure calibration system for an invasive blood pressure monitor. The technical solution is as follows: On the one hand, a dynamic waveform blood pressure calibration system for an invasive blood pressure monitor is provided, the system comprising: The pressure profile module acquires the stroke value of the micro-motor plunger pump, the opening value of the proportional valve, and the output pressure value of the two-stage gas-liquid isolation chamber. It compares the basic pressure setpoint with the output pressure value and generates a pressure profile calibration table. Based on the pressure range calibration table, the segment recognition module collects the standard input waveform pressure value, systolic peak pressure, diastolic trough pressure, and dicrotic notch pressure, and filters the sampling segment after the steep rise edge and the sampling segment adjacent to the notch to obtain a set of resonance candidate segments; The fingerprint registration module calls the set of resonance candidate segments to obtain the catheter resonance frequency position, oscillation start pressure, oscillation end pressure, oscillation duration, and oscillation peak amplitude. It compares the time interval between adjacent oscillation peaks with the change in the amplitude of adjacent oscillation peaks, determines the number of repetitions of the frequency position, the continuous decreasing state of the oscillation peak, the correspondence between the start and end pressure points and the pressure range in the pressure range calibration table, and generates a graded resonance fingerprint table. The waveform verification module uses the graded resonance fingerprint table to call the frequency value, actuation duration value, and output pressure value of the piezoelectric ceramic high-frequency actuator and the dual-stage gas-liquid isolation chamber, compares the output pressure value with the standard blood pressure waveform pressure sampling value, and generates a standard input waveform identifier. Based on the standard input waveform identifier, the compensation and data entry module collects the output pressure value, standard input intensity value, tested output intensity value, standard input phase value, and tested output phase value of the invasive blood pressure monitor, calculates the intensity difference and phase time difference, and establishes a dynamic calibration compensation table.

[0009] As a further embodiment of the present invention, the pressure range calibration table includes a range number, a set pressure level, an output deviation limit, a stable pressure range mark, and a control combination index; the resonance candidate segment set includes a segment number, a waveform stage label, boundary sampling points, local energy indicators, and a disturbance level mark; the range-based resonance fingerprint table includes a fingerprint number, a frequency band assignment code, an attenuation pattern label, a pressure landing area mark, and an oscillation confidence level; the standard input waveform identifier includes a waveform number, an actuation condition code, an input boundary mark, a standard intensity level, and a reference phase label; the dynamic calibration compensation table includes a compensation number, an amplitude correction amount, a timing compensation amount, a frequency response correction label, and an applicable pressure range code.

[0010] As a further aspect of the present invention, the process of screening the sampling segment after the steep rise edge and the sampling segment adjacent to the notch to obtain the resonant candidate segment set includes comparing the pressure change and sampling time interval between adjacent standard blood pressure waveform pressure sampling values, and determining that when the pressure change direction of three consecutive standard blood pressure waveform pressure sampling values ​​is consistent and the pressure change is greater than 5% of the output pressure value in the corresponding pressure range calibration table, the corresponding sampling interval is marked as the sampling segment after the steep rise edge.

[0011] As a further aspect of the present invention, the pressure difference between the dicrotic notch pressure and the pressure sampling value of the adjacent standard blood pressure waveform is less than the notch determination threshold. The notch determination threshold is determined based on the output deviation limit in the corresponding pressure range calibration table, the local fluctuation amplitude of the standard blood pressure waveform pressure sampling value, and the pressure change state within the sampling period. When the pressure difference is less than the notch determination threshold and the adjacent sampling time interval is within the preset sampling period range, the corresponding sampling interval is marked as the notch adjacent sampling segment. The pressure difference between the dicrotic notch pressure and the adjacent standard blood pressure waveform pressure sample value is less than the notch determination threshold. The notch determination threshold is determined based on the output deviation limit in the pressure range calibration table and the local fluctuation amplitude of the standard blood pressure waveform pressure sample value, and is calculated using the following formula: T n =αP e +βV l in: T n The threshold for determining the notch; P e This is the output deviation limit; V l This refers to the local fluctuation amplitude; α and β are weighting coefficients; α + β = 1.

[0012] As a further aspect of the present invention, the pressure recording module includes: The stroke acquisition submodule acquires the stroke value of the micro motor plunger pump, the opening value of the proportional valve, and the output pressure value of the two-stage gas-liquid isolation chamber. It detects the sampling time corresponding to the stroke value and the sampling time corresponding to the opening value, calculates the correspondence between the output pressure change value and the pressure sampling period, determines the matching status between the output pressure change state and the sampling period, and generates a pressure change sequence. The pressure offset comparison submodule collects the basic pressure setpoint and the output pressure value of the dual-stage gas-liquid isolation chamber according to the pressure change sequence, compares the correspondence between the basic pressure setpoint and the output pressure value, calculates the correlation between the pressure difference and the pressure sampling period, determines the distribution state of the pressure interval corresponding to the pressure difference, and obtains the pressure offset interval. Based on the pressure offset range and the pressure change sequence, the gear mapping submodule determines the gear number, sets the pressure level, output deviation limit, pressure stabilization range mark and control combination index in the pressure gear calibration table; The gear number is determined based on the pressure range boundary where the pressure difference between the base pressure set value and the output pressure value of the dual-stage gas-liquid isolation chamber falls. The control combination index is used to characterize the correlation between the stroke value of the micro-motor plunger pump, the opening value of the proportional valve, and the corresponding sampling time.

[0013] As a further aspect of the present invention, the fragment recognition module includes: Based on the pressure range calibration table, the waveform acquisition submodule acquires standard blood pressure waveform pressure sampling values, systolic peak pressure, diastolic trough pressure, and dicrotic notch pressure, detects the corresponding sampling time and pressure range number of the pressure sampling value, calculates the pressure change value between adjacent sampling times, determines the correspondence between the pressure change value and the waveform period, and generates a waveform change sequence. The peak-valley positioning submodule obtains the time corresponding to the peak pressure of systolic blood pressure, the time corresponding to the trough pressure of diastolic blood pressure, and the time corresponding to the pressure of diabetic notch based on the waveform change sequence. It compares the pressure change direction of adjacent sampling times, calculates the interval state corresponding to the peak pressure and the trough pressure, determines the position corresponding to the diabetic notch pressure and the waveform period, and obtains the notch positioning interval. The segment filtering submodule, based on the notch positioning interval, collects the sampling segment after the steep rise edge and the sampling segment adjacent to the notch, detects the relationship between the pressure value corresponding to the sampling segment and the sampling time, calculates the pressure fluctuation state of adjacent sampling segments, determines the corresponding state of the sampling segment and the pressure level number, and obtains the resonance candidate segment set.

[0014] As a further aspect of the present invention, the fingerprint registration module includes: The frequency extraction submodule calls the set of resonance candidate segments to obtain the duct resonance frequency position, oscillation start pressure, oscillation end pressure, oscillation duration, and oscillation peak amplitude. It detects the correspondence between the sampling time corresponding to the frequency position and the oscillation duration, calculates the change state of the oscillation peak amplitude, determines the state of the interval corresponding to the oscillation start pressure and the oscillation end pressure, and generates an oscillation frequency sequence. The peak comparison submodule collects the time interval between adjacent oscillation peaks and the amplitude change value of adjacent oscillation peaks according to the oscillation frequency point sequence, compares the sampling period state corresponding to the time interval with the waveform state corresponding to the amplitude change, calculates the frequency position repetition state and the oscillation peak change state, determines the continuous change relationship of the oscillation peaks, and obtains the peak correlation interval. The interval mapping submodule calls the pressure interval in the pressure range calibration table based on the peak correlation interval, and determines the landing point and duration of the resonance feature in the corresponding pressure interval according to the catheter resonance frequency position, oscillation start pressure, oscillation end pressure and oscillation duration. It compares the correspondence between the frequency position repetition state and the pressure interval to establish a graded resonance fingerprint table.

[0015] As a further aspect of the present invention, the waveform verification module includes: The frequency call submodule calls the frequency value, actuation duration value, and output pressure value of the piezoelectric ceramic high-frequency actuator and the dual-stage gas-liquid isolation chamber according to the graded resonance fingerprint table. It detects the relationship between the sampling time and actuation duration corresponding to the frequency value, calculates the waveform change state corresponding to the output pressure value, determines the relationship between the output pressure value and the pressure grade number, and generates an actuation waveform sequence. Based on the actuated waveform sequence, the waveform pressure comparison submodule collects the standard blood pressure waveform pressure sampling value and the output pressure value of the dual-stage gas-liquid isolation chamber, compares the corresponding state of the output pressure value with the pressure sampling value, calculates the pressure fluctuation change state at adjacent sampling times, determines the correspondence between the pressure fluctuation state and the waveform period, and obtains the waveform matching interval. The waveform calibration submodule obtains the time corresponding to the output pressure value, the time corresponding to the standard blood pressure waveform pressure sampling value, and the interval corresponding to the actuation duration based on the waveform matching interval. It compares the output pressure waveform with the corresponding state of the standard blood pressure waveform, calculates the periodic correlation state of the pressure waveform, and establishes a standard input waveform identifier.

[0016] As a further aspect of the present invention, the compensation storage module includes: Based on the standard input waveform identifier, the intensity acquisition submodule acquires the output pressure value, standard input intensity value, and tested output intensity value of the invasive blood pressure monitor, detects the state corresponding to the sampling time of the output pressure value and the standard input intensity value, calculates the waveform change state corresponding to the tested output intensity value, determines the correspondence between the intensity change state and the pressure waveform period, and generates an intensity offset sequence. The phase comparison submodule obtains the standard input phase value and the detected output phase value based on the intensity offset sequence, compares the corresponding states of the standard input phase value and the detected output phase value, calculates the correspondence between the phase change state and the waveform period, determines the correlation between the phase change state and the intensity offset state, and obtains the phase offset interval. The compensation table creation submodule, based on the phase offset interval, collects the interval corresponding to the intensity difference, the interval corresponding to the phase time difference, and the interval corresponding to the output pressure value of the invasive blood pressure monitor. It compares the corresponding states of the intensity difference and the phase time difference, calculates the periodic correlation state of the pressure waveform, and establishes a dynamic calibration compensation table.

[0017] As a further aspect of the present invention, the process of screening the sampling segment after the steep rise edge and the sampling segment adjacent to the notch further includes: extracting pressure peak points and pressure valley points between the sampling segment after the steep rise edge and the sampling segment adjacent to the notch, determining whether the pressure peak points and pressure valley points are continuously and alternately distributed, and determining the oscillation continuity state of the sampling interval based on the difference between adjacent pressure peak intervals, the sampling period stability condition, and the output deviation limit in the corresponding pressure range calibration table; when the time interval between adjacent pressure peaks meets the preset stability condition, and the amplitude change of adjacent pressure peaks meets the preset attenuation condition, the corresponding sampling interval is marked as a resonance candidate segment and written into the resonance candidate segment set.

[0018] This invention also provides a method for calibrating dynamic waveform blood pressure in an invasive blood pressure monitor applied to the above-mentioned system. The method includes the following steps: a pressure profile establishment step, which involves acquiring the stroke value of the micro-motor plunger pump, the opening value of the proportional valve, and the output pressure value of the two-stage gas-liquid isolation chamber, comparing the baseline pressure setting value with the output pressure value, and generating a pressure profile calibration table; a segment identification step, which involves calling the pressure profile calibration table, collecting the standard input waveform pressure value, the peak systolic pressure, and the dicrotic notch pressure, filtering the sampling segment after the steep rise edge and the sampling segment adjacent to the notch, and generating a set of resonance candidate segments; a fingerprint registration step, which involves calling the set of resonance candidate segments, acquiring the catheter resonance frequency, the oscillation start and end pressure, the oscillation duration, and the oscillation peak amplitude, comparing the time interval and amplitude change of adjacent oscillation peaks, and generating a profiled resonance fingerprint table; a waveform verification step, which involves calling the profiled resonance fingerprint table and the frequency value of the piezoelectric ceramic high-frequency actuator, comparing the output pressure value with the blood pressure waveform pressure sampling value, and generating a standard input waveform identifier; and a compensation storage step, which involves calling the standard input waveform identifier, calculating the intensity difference and the phase time difference, and establishing a dynamic calibration compensation table.

[0019] Compared with the prior art, the advantages and positive effects of the present invention are as follows: 1. This invention achieves dynamic waveform segmented calibration, solving the problem that static calibration cannot reflect dynamic waveform errors. The invention uses a segment recognition module to screen for resonant candidate segments after the rising edge and near the dicrotic notch. It compares the pressure change and sampling time interval between adjacent standard blood pressure waveform pressure sampling values. When three consecutive sampling values ​​show consistent pressure change directions and the change is greater than a preset threshold, the corresponding sampling interval is marked as the sampling segment after the rising edge, preserving key transient information in the dynamic waveform. Simultaneously, a pressure mapping module establishes a segmented mapping relationship between pump stroke, valve opening, and isolation chamber output pressure, enabling calibration to cover the dynamic response differences between high systolic blood pressure and low diastolic blood pressure segments. Example data shows that after adopting this solution, the waveform reconstruction error decreased from 3.9 mmHg to 1.1 mmHg, significantly improving dynamic response consistency.

[0020] 2. This invention achieves graded resonance feature tracking and localization, solving the problem of resonance features not being tracked by grade. The invention uses a fingerprint registration module to access a set of candidate resonance segments, obtaining the catheter resonance frequency position, oscillation initiation pressure, oscillation termination pressure, oscillation duration, and oscillation peak amplitude. It compares the time interval and amplitude changes between adjacent oscillation peaks, determines the correspondence between the frequency position repetition count, the continuous decrease in oscillation peaks, and the initiation and termination pressure points with the pressure ranges in the pressure range calibration table, and generates a graded resonance fingerprint table. By associating resonance features with pressure ranges, the invention can accurately pinpoint whether the deviation originates from the high-pressure or low-pressure segment. Example data shows that the accuracy of range matching has increased from 84% to 97%, and waveform distortion and notch distortion problems under different pressure ranges can be effectively traced.

[0021] 3. Achieving joint compensation for intensity and phase, solving the problem of uncompensated intensity and phase errors. This invention uses a compensation database module based on standard input waveform identifiers to collect the output pressure value of the invasive blood pressure monitor, standard input intensity value, tested output intensity value, standard input phase value, and tested output phase value. It calculates the intensity difference and phase time difference to establish a dynamic calibration compensation table. Simultaneously, the waveform verification module calls a graded resonance fingerprint table and the frequency value of the piezoelectric ceramic high-frequency actuator to compare the output pressure value with the standard blood pressure waveform pressure sampling value. This ensures that the calibration results simultaneously cover the coupling relationship between amplitude deviation, frequency response characteristics, and phase hysteresis. Example data shows that the phase compensation accuracy increased from 81% to 96%, and waveform peak overshoot and fallback hysteresis were effectively suppressed. Attached Figure Description

[0022] 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.

[0023] Figure 1 This is a system flowchart of the present invention; Figure 2 This is a system block diagram of the present invention; Figure 3 This is a flowchart of the pressure data recording module in this invention; Figure 4 This is a flowchart of the segment recognition module in this invention; Figure 5 This is a flowchart of the fingerprint registration module in this invention; Figure 6 This is a flowchart of the waveform verification module in this invention; Figure 7 This is a flowchart of the compensation data entry module in this invention. Detailed Implementation

[0024] The technical solution of the present invention will now be described with reference to the accompanying drawings.

[0025] To make the technical problems, technical solutions and advantages of the present invention clearer, a detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

[0026] Example 1 This invention provides a dynamic waveform blood pressure calibration system for invasive blood pressure monitors, such as... Figure 1-2 The diagram shown illustrates a dynamic waveform blood pressure calibration system for an invasive blood pressure monitor. The system includes: The pressure profile module acquires the stroke value of the micro-motor plunger pump, the opening value of the proportional valve, and the output pressure value of the two-stage gas-liquid isolation chamber. It compares the basic pressure setpoint with the output pressure value and generates a pressure profile calibration table. The micro-motor plunger pump is used to generate basic static pressure and low-frequency pressure changes, the proportional valve is used to regulate medium-frequency pulsating pressure, and the piezoelectric ceramic high-frequency actuator is used to generate dicrotic notches, high-frequency thrills, and steep rise-edge pressure details; the micro-motor plunger pump, proportional valve, and piezoelectric ceramic high-frequency actuator are coupled through a two-stage gas-liquid isolation chamber to form a composite dynamic blood pressure waveform. The two-stage gas-liquid isolation chamber includes a damping membrane layer and a main isolation membrane layer. The damping membrane layer is a polyimide film with a microporous array, which is used to dissipate high-frequency turbulent disturbances. The main isolation membrane layer adopts a gradient thickness silicone film structure. Its central region is coupled with a piezoelectric ceramic high-frequency actuator to transmit high-frequency pressure details, and the outer region is used to receive the proportional valve modulation pressure to improve dynamic waveform stability and frequency response consistency. The system also includes a physiological noise simulation unit, which generates respiratory fluctuations, motion artifacts and vascular pulsation noise, and dynamically adjusts the noise intensity according to the amplitude ratio of the main waveform to construct a dynamic clinical interference environment. The physiological noise simulation unit uses an FPGA hardware acceleration structure to realize the real-time synthesis of respiratory fluctuations, motion artifacts and catheter jitter signals, with noise superposition delay of less than 50 microseconds; The segment recognition module, based on the pressure range calibration table, collects standard input waveform pressure values, systolic peak pressure, diastolic trough pressure, and dicrotic notch pressure, and filters the sampling segments after the steep rise edge and the sampling segments adjacent to the notch to obtain a set of resonance candidate segments; The fingerprint registration module calls the resonance candidate fragment set to obtain the duct resonance frequency position, oscillation start pressure, oscillation end pressure, oscillation duration, and oscillation peak amplitude. It compares the time interval between adjacent oscillation peaks with the change in the amplitude of adjacent oscillation peaks, determines the number of repetitions of the frequency position, the continuous decreasing state of the oscillation peak, the correspondence between the start and end pressure points and the pressure range in the pressure range calibration table, and generates a graded resonance fingerprint table. The system also includes a waveform storage unit, which stores standard blood pressure waveform sets, pathological waveform sets, and custom waveform data. It uses a dynamic loading method to call the corresponding waveform segments on demand, thereby reducing storage resource consumption and improving waveform calling efficiency. The waveform verification module uses the graded resonance fingerprint table to call the frequency value, actuation duration value, and output pressure value of the piezoelectric ceramic high-frequency actuator and the dual-stage gas-liquid isolation chamber, compares the output pressure value with the standard blood pressure waveform pressure sampling value, and generates a standard input waveform identifier. The compensation and data entry module is based on the standard input waveform identifier. It collects the output pressure value, standard input intensity value, tested output intensity value, standard input phase value, and tested output phase value of the invasive blood pressure monitor, calculates the intensity difference and phase time difference, and establishes a dynamic calibration compensation table. The damping membrane has a micropore diameter of 10~50μm and a porosity of 5%~15%. The thickness of the central region of the main isolation membrane is 0.3~0.6mm, and the thickness of the edge region is 0.8~1.2mm. The preset pressure change threshold is determined based on the pressure fluctuation range corresponding to the pressure setting, preferably 2 mmHg to 6 mmHg. The "standard input waveform pressure value" in this manual refers to the standard reference pressure waveform data output by the dynamic waveform generation system; "Patient output pressure value" refers to the pressure data output by the invasive blood pressure monitor after collecting the standard reference pressure waveform.

[0027] The pressure range calibration table includes the range number, set pressure level, output deviation limit, pressure stabilization range mark, and control combination index; the resonance candidate segment set includes the segment number, waveform stage label, boundary sampling point, local energy index, and disturbance level mark; the ranged resonance fingerprint table includes the fingerprint number, frequency band assignment code, attenuation pattern label, pressure landing area mark, and oscillation confidence level; the standard input waveform identifier includes the waveform number, actuation condition code, input boundary mark, standard intensity level, and reference phase label; the dynamic calibration compensation table includes the compensation number, amplitude correction amount, timing compensation amount, frequency response correction label, and applicable pressure range code.

[0028] Specifically, such as Figure 2 , 3 As shown, the pressure data entry module includes: The stroke acquisition submodule acquires the stroke value of the micro motor plunger pump, the opening value of the proportional valve, and the output pressure value of the two-stage gas-liquid isolation chamber. It detects the sampling time corresponding to the stroke value and the sampling time corresponding to the opening value, calculates the correspondence between the output pressure change value and the pressure sampling period, determines the matching status between the output pressure change state and the sampling period, and generates a pressure change sequence. When collecting the stroke value of the micro-motor plunger pump, the plunger displacement sensor outputs displacement data frames at a sampling period of 5 milliseconds. The displacement data frames are stored in a 16-bit pressure control register format, and the acquisition range is set to 0 mm to 12 mm. In this embodiment, the real-time displacement value of the plunger is 6.4 mm. The proportional valve opening sensor synchronously outputs the opening percentage corresponding to the valve core deflection angle, and the current acquisition value is 42%. The pressure sensor of the two-stage gas-liquid isolation chamber outputs the pressure sampling value using a piezoresistive pressure chip, and the current sampling value is 118 mmHg. A unified time tag is attached to the three sets of data. The current unified sampling time is 15:23:18:205. Subsequently, a sampling time matching action is performed. When the time difference between the plunger displacement data frame and the proportional valve opening data frame is less than 2 milliseconds, it is determined to be a valid synchronization frame. In this embodiment, the time difference is 1 millisecond, which meets the synchronization condition. Subsequently, the output pressure values ​​from three consecutive pressure sampling cycles are retrieved. The pressure value of the previous cycle is 112 mmHg, the pressure value of the current cycle is 118 mmHg, and the pressure value of the next cycle is 121 mmHg. A difference operation is performed on the consecutive pressure values ​​to obtain a current pressure change of 6 mmHg and a next cycle change of 3 mmHg, and the pressure change value is written to the pressure change buffer. The pressure sampling cycle numbers are registered in a sequentially increasing manner. In this embodiment, the current cycle number is 2054. Then, the pressure change value is correlated with the pressure sampling cycle. When the pressure change value is within the range of 2 mmHg to 8 mmHg for two consecutive sampling cycles, the pressure change state is determined to be in a stable pressure increase state. In this embodiment, both 6 mmHg and 3 mmHg satisfy the range condition, therefore a stable pressure increase marker is generated. The plunger displacement value of 6.4 mm, the proportional valve opening of 42%, the output pressure of 118 mmHg, and the stable pressure increase marker are combined and registered as data frame number 2054 of the pressure change sequence. Table 1 Pressure Change and Pressure Range Calibration Data ; As shown in Table 1, the real-time plunger displacement values ​​in sampling periods 2054 to 2058 remained within the range of 6 mm to 7 mm, the proportional valve opening value remained within the range of 41% to 44%, and the pressure change value remained within the range of 2 mmHg to 6 mmHg. Therefore, all of these were recorded as stable pressure-boosting data frames. In the experiment, compared with the traditional time-synchronized acquisition method, 11 sets of data frames out of 100 consecutive pressure samples had a deviation exceeding 5 milliseconds. After synchronization processing in this embodiment, the number of sets with deviations exceeding the limit was reduced to 2 sets, and the sampling consistency rate was improved from 89% to 98%.

[0029] The pressure offset comparison submodule collects the basic pressure setpoint and the output pressure value of the dual-stage gas-liquid isolation chamber according to the pressure change sequence, compares the correspondence between the basic pressure setpoint and the output pressure value, calculates the correlation between the pressure difference and the pressure sampling period, determines the distribution state of the pressure interval corresponding to the pressure difference, and obtains the pressure offset interval. After calling the pressure change sequences 2054 to 2058 in Table 1, the baseline pressure setting value is read. This baseline pressure setting value is pre-input via the pressure control panel; the current setting value is 120 mmHg. Then, the output pressure value of the dual-stage gas-liquid isolation chamber (118 mmHg) is read, and a pressure correspondence comparison is performed. When the difference between the output pressure value and the baseline pressure setting value is less than 10 mmHg, it is registered as a normal offset interval. In this embodiment, the pressure difference is 2 mmHg, satisfying the normal offset condition. The pressure difference values ​​for five consecutive sampling periods in Table 1 are read: 2 mmHg, 3 mmHg, 2 mmHg, 4 mmHg, and 3 mmHg, respectively. Interval distribution statistics are then performed on these pressure difference values. In this embodiment, the pressure difference values ​​are concentrated in the 0 mmHg to 5 mmHg range, accounting for 100%, and therefore are registered as a first-level pressure offset interval. The pressure sampling period association status is processed using a continuous numbering binding method. The current sampling period numbers 2054 to 2058 are all written into the first-level pressure offset interval index table. A pressure fluctuation threshold, determined through statistical analysis of 200 sets of standard waveforms, is set at 8 mmHg. When the pressure difference variation within a continuous sampling period is less than 8 mmHg, it is registered as a stable pressure offset state. In this embodiment, the maximum variation is 2 mmHg, thus triggering a stable offset flag. The statistical results of the pressure difference are associated with the sampling period number and stored to form a pressure offset interval registration record. In the experiment, the waveform offset identification error after using the first-level pressure offset interval was 1.8 mmHg, while the error reached 5.6 mmHg without interval association, resulting in a 67% improvement in pressure offset stability.

[0030] The gear mapping submodule determines the gear number, sets the pressure level, output deviation limit, pressure stabilization range mark and control combination index in the pressure gear calibration table based on the pressure offset range and pressure change sequence. The gear number is determined based on the pressure range boundary where the pressure difference between the basic pressure setpoint and the output pressure of the dual-stage gas-liquid isolation chamber falls. The control combination index is used to characterize the correlation between the stroke value of the micro-motor plunger pump, the opening value of the proportional valve, and the corresponding sampling time. After calling the first-level pressure offset range in Table 1, the basic pressure setting range and output pressure segment range are read. The basic pressure setting range is divided in 20 mmHg intervals. The current basic pressure setting value of 120 mmHg corresponds to the 100 mmHg to 120 mmHg segment. The output pressure segment range is divided according to the actual sampled value. The current output pressure of 118 mmHg corresponds to the 110 mmHg to 130 mmHg segment. The pressure range boundary values ​​are read. The current first-level pressure range boundary is set to 0 mmHg to 5 mmHg, and the second-level pressure range boundary is set to 5 mmHg to 10 mmHg. In this embodiment, the pressure difference in Table 1 is 2 mmHg, so the current data is determined to belong to the first-level pressure range. Then, sampling period number 2054 is called, and the correspondence between the pressure range number and the sampling period number is established. The current first-level pressure range corresponds to sampling periods 2054 to 2058. The pressure range calibration table registration process is performed, recording the baseline pressure range (100 mmHg to 120 mmHg), the output pressure range (110 mmHg to 130 mmHg), the primary pressure range number, and the sampling cycle numbers (2054 to 2058) into the calibration table. To avoid cross-classification of pressure ranges, boundary verification is performed on 20 consecutive sets of pressure data. When the pressure difference exceeds the current range boundary three times consecutively, a range switching action is triggered. In this embodiment, all 20 consecutive sets of data are below 5 mmHg, thus maintaining the primary pressure range status. In the experiment, range mapping verification was performed on 500 sets of blood pressure waveforms. After using the pressure range calibration table, the range matching accuracy reached 97%, while the accuracy of the traditional fixed threshold method was 84%, and the number of misclassified ranges was reduced from 80 sets to 15 sets.

[0031] Specifically, such as Figure 2 , 4 As shown, the fragment recognition module includes: The waveform acquisition submodule is based on the pressure range calibration table. It acquires standard blood pressure waveform pressure sampling values, systolic peak pressure, diastolic trough pressure, and dicrotic notch pressure. It detects the corresponding state of the sampling time and pressure range number of the pressure sampling value, calculates the pressure change value between adjacent sampling times, determines the correspondence between the pressure change value and the waveform period, and generates a waveform change sequence. After retrieving the primary pressure level number from the pressure level calibration table, the system reads the pressure sample values ​​from the standard blood pressure waveform database. The waveform data in the database is output by the clinical standard calibrator, with a sampling frequency set to 200 Hz. The currently acquired standard pressure sample value is 118 mmHg, the peak systolic pressure is 126 mmHg, the trough diastolic pressure is 78 mmHg, and the dicrotic notch pressure is 92 mmHg. A sampling time label is attached to the pressure sample value; the current sampling time is 15:23:18:240, and the corresponding primary pressure level number is read. Subsequently, the pressure change value is calculated for adjacent sampling times. The previous pressure value was 113 mmHg, the current pressure value is 118 mmHg, therefore the current pressure change value is 5 mmHg; the next pressure value is 122 mmHg, therefore the next change value is 4 mmHg. The continuously changing values ​​were compared with the standard waveform period, which was obtained from the statistical analysis of 300 sets of clinical waveforms. The normal period range was set to 0.6 seconds to 1.1 seconds. In this embodiment, the current period was 0.82 seconds, and therefore it was registered as a normal waveform period. A waveform change sequence was then generated. Sequence number 2054 recorded the 118 mmHg sampling value, the first-level pressure level number, the 0.82-second waveform period, and the 5 mmHg pressure change value. Filtering was performed on 50 consecutive sets of sampling values ​​using a 3-point median rejection method. When a single point fluctuation exceeded the mean of 12 mmHg, it was determined to be an abnormal pulse. In this embodiment, there was one abnormal point, which was rejected. Experimental results showed that after filtering, the waveform slope error decreased from 7.4% to 2.1%, and the waveform period stability improved by 31%.

[0032] The peak-valley positioning submodule obtains the time corresponding to the peak pressure of systolic blood pressure, the time corresponding to the trough pressure of diastolic blood pressure, and the time corresponding to the pressure of diabetic notch based on the waveform change sequence. It compares the pressure change direction of adjacent sampling times, calculates the interval state corresponding to the peak pressure and the trough pressure, determines the position corresponding to the diabetic notch pressure and the waveform period, and obtains the notch positioning interval. After reading the waveform change sequence, the peak systolic pressure was extracted at 15:23:18.410 ms, the trough diastolic pressure at 15:23:18.920 ms, and the dicrotic notch pressure at 15:23:18.635 ms. Then, the pressure change direction of adjacent sampling times was retrieved. When the pressure value increased for two consecutive sampling periods, it was recorded as an upward direction; when the pressure value decreased for two consecutive sampling periods, it was recorded as a downward direction. In this embodiment, the increase from 118 mmHg to 126 mmHg was recorded as an upward direction, and the decrease from 126 mmHg to 92 mmHg was recorded as a downward direction. The time interval between the peak and trough pressures was statistically analyzed. The current time interval was 510 ms, and it was compared with the preset normal range of 400 ms to 650 ms. The current value met the normal range, therefore a peak-trough stability marker was added. The position of the dicrotic notch pressure within the waveform period was then located. The current notch position is 225 milliseconds from the peak, accounting for 27% of the overall waveform period of 0.82 seconds. The preset notch position range is 20% to 35%, therefore it is registered as a valid notch position. A notch positioning range is generated, currently set to 15:23:18.620 milliseconds to 15:23:18.650 milliseconds. Table 2. Waveform notch localization and resonance segment selection data. ; As shown in Table 2, the peak-valley time intervals in waveform sequences 2054 to 2058 all remained between 505 and 515 milliseconds, with the notch localization interval concentrated around 620 milliseconds. Therefore, all of these were registered as valid notch localization data. In the experiment, peak-valley localization tests were performed on 300 sets of standard blood pressure waveforms. In this embodiment, the average notch localization error was 6 milliseconds, while the error of the traditional fixed peak detection method was 19 milliseconds, representing a 68% reduction in localization error.

[0033] The segment selection submodule is based on the notch positioning interval, collects the sampling segment after the steep rise edge and the sampling segment adjacent to the notch, detects the relationship between the pressure value corresponding to the sampling segment and the sampling time, calculates the pressure fluctuation state of adjacent sampling segments, determines the corresponding state of the sampling segment and the pressure level number, and obtains the resonance candidate segment set. After calling the notch location interval in Table 2, the sampling segment after the rising edge and the sampling segment adjacent to the notch are read. The starting pressure value of the sampling segment after the rising edge is 102 mmHg, and the ending pressure value is 126 mmHg, corresponding to the time interval from 15:23:18.300 ms to 15:23:18.410 ms; the starting pressure value of the sampling segment adjacent to the notch is 96 mmHg, and the ending pressure value is 92 mmHg, corresponding to the time interval from 15:23:18.620 ms to 15:23:18.650 ms. The pressure values ​​within the sampling segments are mapped and registered with the sampling time, and then the pressure fluctuation status of adjacent sampling segments is statistically analyzed. In this embodiment, the segment fluctuation value in Table 2 is 2.1 mmHg per 10 ms, and the average fluctuation value of the sampling segment adjacent to the notch decreases by 1.3 mmHg per 10 ms. Subsequently, the primary pressure level number is read, and the association between the sampling segment and the pressure level number is executed. When the fluctuation value of the sampling segment is within the allowable fluctuation range of the primary pressure level (0.5 mmHg to 3 mmHg per 10 milliseconds), it is registered as a valid resonance candidate segment. In this embodiment, both 2.1 mmHg and 1.3 mmHg meet the condition, and therefore are added to the resonance candidate segment set. Repeatability verification is performed on 30 consecutive sets of candidate segments. When the consistency rate of fluctuation direction between adjacent segments exceeds 85%, the segment is determined to be stable. In this embodiment, the consistency rate reaches 93%, and therefore it is registered as a stable resonance candidate segment. Experimental data show that the segment selection accuracy after pressure level association reaches 95%, while the accuracy rate of the traditional time window interception method is 81%, and the number of incorrectly selected segments is reduced from 57 to 14.

[0034] Specifically, such as Figure 2 , 5 As shown, the fingerprint registration module includes: The frequency extraction submodule calls the resonance candidate fragment set to obtain the duct resonance frequency position, oscillation start pressure, oscillation end pressure, oscillation duration, and oscillation peak amplitude. It detects the correspondence between the sampling time corresponding to the frequency position and the oscillation duration, calculates the oscillation peak amplitude change status, determines the interval status corresponding to the oscillation start pressure and oscillation end pressure, and generates an oscillation frequency sequence. The resonant frequency position of the catheter was obtained by conversion through continuous peak intervals. In this embodiment, the continuous peak interval is 42 milliseconds, corresponding to a resonant frequency position of 23.8 Hz. The oscillation start pressure is 96 mmHg, the oscillation end pressure is 121 mmHg, the oscillation duration is 280 milliseconds, and the oscillation peak amplitude is 11 mmHg. The sampling time corresponding to the frequency position is registered; the current sampling time is 15:23:18.640, and the oscillation duration is bound to the sampling time. Subsequently, five consecutive peak amplitude data points are read: 11 mmHg, 10 mmHg, 12 mmHg, 11 mmHg, and 10 mmHg, and change status statistics are performed. The current maximum fluctuation value is 2 mmHg, which is lower than the preset peak stability threshold of 5 mmHg, therefore it is registered as a stable oscillation state. An interval judgment is performed on the oscillation start pressure and oscillation end pressure. When the pressure span is between 20 mmHg and 35 mmHg, it is registered as a valid oscillation interval. In this embodiment, the pressure span is 25 mmHg, therefore a valid oscillation frequency sequence is generated. In the experiment, the resonance identification accuracy using the oscillation frequency sequence reached 96%, while the accuracy of the traditional single-peak identification method was 83%, and the frequency misidentification rate decreased by 13%.

[0035] The peak comparison submodule collects the time interval between adjacent oscillation peaks and the amplitude change value of adjacent oscillation peaks according to the oscillation frequency point sequence. It compares the sampling period state corresponding to the time interval with the waveform state corresponding to the amplitude change, calculates the frequency position repetition state and the oscillation peak change state, determines the continuous change relationship of the oscillation peak, and obtains the peak correlation interval. After reading the oscillation frequency sequence, the time interval between adjacent oscillation peaks and the amplitude change value of adjacent oscillation peaks are retrieved. The current time intervals between adjacent oscillation peaks are 42 milliseconds, 44 milliseconds, and 43 milliseconds, respectively, and the amplitude changes of adjacent oscillation peaks are 1 mmHg, 2 mmHg, and 1 mmHg, respectively. The sampling period states corresponding to the time intervals are compared. When the time interval deviation is less than 5 milliseconds, it is registered as a stable sampling period. In this embodiment, the maximum deviation is 2 milliseconds, thus satisfying the stability condition. Subsequently, an association judgment is performed between the amplitude change value and the waveform state. When the continuous peak amplitude change is less than 3 mmHg, it is registered as a continuous stable oscillation state. In this embodiment, all change values ​​are less than 3 mmHg, thus generating a continuous oscillation marker. The frequency position repetition state is statistically analyzed. The current three consecutive frequency positions are 23.8 Hz, 24.1 Hz, and 23.9 Hz, with a frequency fluctuation range of 0.3 Hz. The preset allowable repetition frequency range is 1 Hz, therefore, it is registered as a frequency position repetition state. Subsequently, a peak correlation interval was generated, currently covering the period from 15:23:18.640 ms to 15:23:18.760 ms. In the experiment, the resonance continuity identification accuracy after adopting the peak correlation interval reached 94%, while the accuracy of the traditional fixed frequency detection method was 79%, and the false negative rate of continuous oscillation identification decreased by 15%.

[0036] The interval mapping submodule is based on the peak correlation interval, calls the pressure interval in the pressure range calibration table, and determines the landing point and duration of the resonance feature in the corresponding pressure interval according to the position of the duct resonance frequency, the oscillation start pressure, the oscillation end pressure and the oscillation duration. It compares the correspondence between the repetition state of the frequency position and the pressure interval to establish a graded resonance fingerprint table. After calling the peak correlation interval, the primary pressure interval, the corresponding frequency position interval, and the corresponding oscillation start pressure interval are read from the pressure range calibration table. The current primary pressure interval is 100 mmHg to 120 mmHg, the corresponding frequency position interval is 23 Hz to 25 Hz, and the corresponding oscillation start pressure interval is 90 mmHg to 100 mmHg. A correlation comparison is performed between the frequency position repetition state and the pressure interval. In this embodiment, the 23.8 Hz frequency position falls within the 23 Hz to 25 Hz interval and corresponds to the primary pressure interval, therefore it is registered as a primary resonance fingerprint. Subsequently, the oscillation termination pressure of 121 mmHg and the oscillation duration of 280 ms are read, and correlation state statistics are performed. When the oscillation termination pressure is between 120 mmHg and 130 mmHg, and the oscillation duration is between 250 ms and 320 ms, it is registered as a stable resonance state. In this embodiment, both parameters meet the conditions, therefore they are written into the graded resonance fingerprint table. Repeatability verification was performed on 100 consecutive sets of fingerprint data. The current repeatability matching rate reached 96%, while the preset stable matching threshold was 90%. Therefore, the first-level resonant fingerprint state was maintained. Table 3 Dynamic calibration compensation parameter data table ; As shown in Table 3, the resonant frequencies are all concentrated in the 23 Hz to 25 Hz range, and the oscillation duration remains in the 275 ms to 285 ms range. Therefore, all of them are registered as first-order stable resonant fingerprints. In the experiment, after using the graded resonant fingerprint table, the accuracy of resonant interval classification reached 97%, while the accuracy of the traditional single interval classification method was 82%, and the number of misclassifications decreased from 54 groups to 11 groups.

[0037] Specifically, such as Figure 2 , 6 As shown, the waveform verification module includes: The frequency call submodule calls the frequency value, actuation duration value, and output pressure value of the piezoelectric ceramic high-frequency actuator and the dual-stage gas-liquid isolation chamber according to the graded resonance fingerprint table. It detects the relationship between the sampling time and actuation duration corresponding to the frequency value, calculates the waveform change state corresponding to the output pressure value, determines the relationship between the output pressure value and the pressure grade number, and generates an actuation waveform sequence. After reading the graded resonance fingerprint data in Table 3, the frequency value of the piezoelectric ceramic high-frequency actuator (24 Hz), the actuation duration (280 ms), and the output pressure value of the dual-stage gas-liquid isolation chamber (118 mmHg) are retrieved. A sampling time label is attached to the frequency value; the current sampling time is 15:23:19:020 ms. The actuation duration is then bound and stored with the frequency value. Subsequently, three consecutive output pressure sampling values ​​are read: 118 mmHg, 121 mmHg, and 119 mmHg, and waveform change status statistics are performed. The current pressure fluctuation range is 3 mmHg, which is below the first-level pressure range's allowable fluctuation threshold of 6 mmHg; therefore, it is registered as a stable actuation waveform state. The output pressure value is associated with the pressure range number. When the output pressure value is within the first-level pressure range of 100 mmHg to 120 mmHg, it is registered as a first-level actuation state. In this embodiment, 118 mmHg meets the condition, therefore, an actuation waveform sequence is generated. In the experiment, 200 sets of actuation waveforms were tested. The stability rate of the actuation waveform in this embodiment reached 95%, while the stability rate of the traditional fixed frequency drive method was 80%. The number of pressure fluctuations exceeding the limit was reduced from 40 sets to 10 sets.

[0038] The waveform pressure comparison submodule is based on the actuated waveform sequence. It collects the standard blood pressure waveform pressure sampling value and the output pressure value of the dual-stage gas-liquid isolation chamber, compares the corresponding state of the output pressure value with the pressure sampling value, calculates the pressure fluctuation change state at adjacent sampling times, determines the correspondence between the pressure fluctuation state and the waveform period, and obtains the waveform matching interval. After invoking the actuation waveform sequence, the standard blood pressure waveform pressure sample value of 120 mmHg and the output pressure value of the dual-stage gas-liquid isolation chamber of 118 mmHg are read, and a corresponding state comparison is performed. The current pressure difference is 2 mmHg, which is lower than the waveform matching threshold of 5 mmHg, so it is registered as a valid matching state. Subsequently, the pressure values ​​of 116 mmHg, 118 mmHg, and 120 mmHg at adjacent sampling times are read, and the pressure fluctuation changes are counted. The current fluctuation value is 2 mmHg. The output pressure value and waveform period value data in Table 3 are retrieved. The five consecutive sets of output pressure values ​​in Table 3 are 119 mmHg, 120 mmHg, 118 mmHg, 121 mmHg, and 119 mmHg, with corresponding waveform period values ​​of 0.81 seconds, 0.82 seconds, 0.80 seconds, 0.81 seconds, and 0.81 seconds, respectively. The current maximum period deviation is 0.02 seconds, which is lower than the preset period stability threshold of 0.05 seconds, so all are registered as stable matching waveforms. A waveform matching interval was then generated, currently covering the period from 15:23:19.020 ms to 15:23:19.830 ms. In the experiment, the standard waveform alignment error after using the waveform matching interval was 1.6 mmHg, while the error of the traditional single-point comparison method was 4.8 mmHg, representing a 66% reduction in alignment error.

[0039] The waveform calibration submodule obtains the time corresponding to the output pressure value, the time corresponding to the standard blood pressure waveform pressure sampling value, and the interval corresponding to the actuation duration based on the waveform matching interval. It compares the output pressure waveform with the corresponding state of the standard blood pressure waveform, calculates the periodic correlation state of the pressure waveform, and establishes a standard input waveform identifier. After reading the waveform matching interval, the system retrieves the output pressure value corresponding to time 15:23:19.200, the standard blood pressure waveform pressure sampling value corresponding to time 15:23:19.210, and the arousal duration corresponding to the interval 260 to 300 milliseconds. An overlap comparison is performed between the output pressure waveform and the standard blood pressure waveform. The current peak output pressure is 121 mmHg, while the standard waveform peak is 122 mmHg, resulting in a peak deviation of 1 mmHg, which is lower than the preset waveform deviation threshold of 4 mmHg. Therefore, it is registered as a valid calibration state. Subsequently, the system retrieves the resonant frequency positions and output pressure value data from Table 3. The five consecutive resonant frequency positions in Table 3 are 23.8 Hz, 24.1 Hz, 23.9 Hz, 24.0 Hz, and 23.7 Hz, corresponding to output pressure values ​​of 119 mmHg, 120 mmHg, 118 mmHg, 121 mmHg, and 119 mmHg, respectively. The current frequency fluctuation range is 0.4 Hz, and the output pressure fluctuation range is 3 mmHg, thus meeting all stability calibration conditions. A standard input waveform identifier is then established, currently recording the first-level pressure range number, the 24 Hz actuation frequency, the 0.81-second output cycle, and the 121 mmHg output peak value. In the experiment, the waveform cycle matching accuracy in this embodiment reached 98%, while the accuracy of the traditional fixed-cycle matching method was 85%.

[0040] Specifically, such as Figure 2 , 7 As shown, the compensation data entry module includes: The intensity acquisition submodule acquires the output pressure value, standard input intensity value, and tested output intensity value of the invasive blood pressure monitor based on the standard input waveform identifier. It detects the state corresponding to the sampling time of the output pressure value and the standard input intensity value, calculates the waveform change state corresponding to the tested output intensity value, determines the correspondence between the intensity change state and the pressure waveform period, and generates an intensity offset sequence. After calling the standard input waveform identifier, the output pressure value of the invasive blood pressure monitor (119 mmHg), the standard input intensity value (12 mmHg), and the tested output intensity value (10 mmHg) are read. The current intensity difference is 2 mmHg. Subsequently, the tested output intensity values ​​of 10 mmHg, 11 mmHg, 9 mmHg, and 10 mmHg are read over four consecutive waveform cycles, and the waveform change status is statistically analyzed. The current maximum fluctuation value is 2 mmHg, which is lower than the preset intensity stability threshold of 5 mmHg, so it is registered as a stable intensity state. The intensity offset value data in Table 3 is called. The five consecutive sets of intensity offset values ​​in Table 3 are 2 mmHg, 2 mmHg, 3 mmHg, 2 mmHg, and 2 mmHg, corresponding to output pressure values ​​of 119 mmHg, 120 mmHg, 118 mmHg, 121 mmHg, and 119 mmHg, respectively. The current intensity offset values ​​are all within the intensity offset judgment range of 0 mmHg to 3 mmHg, so an intensity offset sequence is generated. In the experiment, the intensity shift recognition error using the intensity shift sequence was 1.3 mmHg, while the error of the traditional fixed gain detection method was 4.1 mmHg.

[0041] The phase comparison submodule obtains the standard input phase value and the detected output phase value based on the intensity offset sequence, compares the corresponding states of the standard input phase value and the detected output phase value, calculates the correspondence between the phase change state and the waveform period, determines the correlation between the phase change state and the intensity offset state, and obtains the phase offset interval. After invoking the intensity offset sequence, the standard input phase value of 0.82 seconds and the detected output phase value of 0.79 seconds are read, and a phase-correspondence state comparison is performed. The current phase difference is 0.03 seconds, which is lower than the preset phase offset threshold of 0.08 seconds, so it is registered as a level one phase offset state. Subsequently, the phase change values ​​of 0.03 seconds, 0.02 seconds, 0.04 seconds, 0.03 seconds, and 0.03 seconds within five consecutive waveform cycles are read, and the waveform cycle correlation state is statistically analyzed. The current maximum fluctuation value is 0.02 seconds, which is lower than the preset cycle fluctuation threshold of 0.05 seconds, so a stable phase state mark is generated. The phase offset values ​​and waveform cycle values ​​in Table 3 are retrieved. Currently, all phase offset values ​​are lower than 0.05 seconds, and all waveform cycle values ​​are in the range of 0.75 seconds to 0.90 seconds, so a phase offset interval is generated. In the experiment, the phase compensation accuracy after using the phase offset interval reached 96%, while the accuracy of the traditional single-phase correction method was 81%.

[0042] The compensation table creation submodule is based on the phase offset interval, collects the interval corresponding to the intensity difference, the interval corresponding to the phase time difference, and the interval corresponding to the output pressure value of the invasive blood pressure monitor, compares the corresponding state of the intensity difference with the phase time difference, calculates the periodic correlation state of the pressure waveform, and establishes a dynamic calibration compensation table. After calling the phase offset interval, the intervals corresponding to the intensity difference (0 mmHg to 3 mmHg), the phase time difference (0 seconds to 0.05 seconds), and the output pressure value of the invasive blood pressure monitor (110 mmHg to 130 mmHg) are read. The current intensity difference is 2 mmHg, and the phase time difference is 0.03 seconds, therefore it is registered as a Level 1 dynamic compensation state. Subsequently, the pressure waveform corresponding to a period of 0.81 seconds is read, and 20 consecutive sets of period data are statistically analyzed. The current period fluctuation range is 0.04 seconds, lower than the preset period stability threshold of 0.06 seconds, therefore a stable period marker is generated. The intensity offset value, phase offset value, and output pressure value data from Table 3 are called. Currently, all intensity offset values ​​fall within the interval corresponding to the intensity difference, and all phase offset values ​​fall within the interval corresponding to the phase time difference. Therefore, a dynamic calibration compensation table is established. The current compensation table records the Level 1 pressure range number, the 24 Hz actuation frequency, the 2 mmHg intensity offset value, the 0.03-second phase offset value, and the 0.81-second pressure period value. In the experiment, the pressure waveform reconstruction error after using a dynamic calibration compensation gauge was 1.1 mmHg, while the error of the traditional fixed compensation gauge method was 3.9 mmHg.

[0043] The specific operation procedures of the pressure filing module, segment recognition module, fingerprint registration module, waveform verification module and compensation storage module mentioned above correspond to the method steps of this invention, and will not be repeated here.

[0044] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A dynamic waveform blood pressure calibration system for an invasive blood pressure monitor, characterized in that, The system includes: The pressure profile module is used to obtain the stroke value of the micro-motor plunger pump, the opening value of the proportional valve, and the output pressure value of the two-stage gas-liquid isolation chamber. It compares the basic pressure setpoint with the output pressure value and generates a pressure profile calibration table. The segment recognition module is used to call the pressure range calibration table, collect standard input waveform pressure values, systolic blood pressure peak pressure, and dicrotic notch pressure, filter the sampling segment after the steep rise edge and the sampling segment adjacent to the notch, and generate a set of resonance candidate segments; The fingerprint registration module is used to call the set of resonance candidate segments, obtain the duct resonance frequency, oscillation start and stop pressure, oscillation peak amplitude, compare the time interval and amplitude change of adjacent oscillation peaks, determine the frequency position repetition number and oscillation peak decreasing state, and generate a graded resonance fingerprint table. The waveform verification module is used to call the graded resonance fingerprint table, the frequency value of the piezoelectric ceramic high-frequency actuator, and the output pressure value of the dual-stage gas-liquid isolation chamber, compare the output pressure value with the blood pressure waveform pressure sampling value, and generate a standard input waveform identifier. The compensation input module is used to call the standard input waveform identifier, obtain the output pressure value, standard input intensity value, tested output intensity value, and input phase value of the invasive blood pressure monitor, calculate the intensity difference and phase time difference, and establish a dynamic calibration compensation table.

2. The dynamic waveform blood pressure calibration system for invasive blood pressure monitors according to claim 1, characterized in that, The pressure range calibration table includes a range number, set pressure level, output deviation limit, stable pressure range mark, and control combination index; the resonance candidate segment set includes a segment number, waveform stage label, boundary sampling point, local energy index, and disturbance level mark; the ranged resonance fingerprint table includes a fingerprint number, frequency band assignment code, attenuation pattern label, pressure landing area mark, and oscillation confidence level; the standard input waveform identifier includes a waveform number, actuation condition code, input boundary mark, standard intensity level, and reference phase label; the dynamic calibration compensation table includes a compensation number, amplitude correction amount, timing compensation amount, frequency response correction label, and applicable pressure range code.

3. The dynamic waveform blood pressure calibration system for invasive blood pressure monitors according to claim 1, characterized in that, The process of screening the sampling segment after the steep rise edge and the sampling segment adjacent to the notch includes: comparing the pressure change and sampling time interval between adjacent standard blood pressure waveform pressure sampling values; when the pressure change direction of three consecutive standard blood pressure waveform pressure sampling values ​​is consistent and the pressure change is greater than a preset pressure change threshold, the corresponding sampling interval is marked as the sampling segment after the steep rise edge; when the pressure difference between the dicrotic notch pressure and the pressure sampling value of the adjacent standard blood pressure waveform is less than the notch determination threshold and the adjacent sampling time interval is within a preset sampling period, the corresponding sampling interval is marked as the sampling segment adjacent to the notch; the calculation formula for the notch determination threshold is T. n =αP e +βV l in: T n The threshold for determining the notch; P e This is the output deviation limit; V l This refers to the local fluctuation amplitude; α and β are weighting coefficients.

4. The dynamic waveform blood pressure calibration system for an invasive blood pressure monitor according to claim 1, characterized in that, The pressure data acquisition module includes a stroke acquisition submodule, a pressure offset comparison submodule, and a gear mapping submodule. The stroke acquisition submodule is used to synchronously acquire the stroke value of the micro-motor plunger pump, the opening value of the proportional valve, and the output pressure value of the two-stage gas-liquid isolation chamber, match the sampling time of each parameter, calculate the pressure change, determine the pressure change state, and generate a pressure change sequence. The pressure offset comparison submodule is used to retrieve the basic pressure setpoint, compare it with the output pressure value of the two-stage gas-liquid isolation chamber, calculate the pressure difference, and statistically analyze the pressure interval distribution to obtain the pressure offset interval. The pressure level mapping submodule is used to combine the pressure change sequence and pressure offset range to divide the pressure levels and configure the corresponding parameters, and determine the pressure level calibration table.

5. The dynamic waveform blood pressure calibration system for an invasive blood pressure monitor according to claim 1, characterized in that, The segment recognition module includes a waveform acquisition submodule, a peak-valley location submodule, and a segment filtering submodule. The waveform acquisition submodule calls the pressure level calibration table to acquire various pressure parameters of the standard input waveform, calculates the pressure change based on the sampling time, associates the waveform periodic characteristics, and generates a waveform change sequence. The peak-valley location submodule, based on the waveform change sequence, locates the time-series nodes corresponding to the systolic blood pressure peak, diastolic blood pressure trough, and dicrotic notch, determines the direction of pressure change, and delineates the notch location interval. The segment filtering submodule, based on the notch location interval, extracts the sampling segment after the steep rise edge and the sampling segment adjacent to the notch, verifies the matching relationship between the sampling segment and the pressure level, filters valid oscillation segments, and obtains a set of resonance candidate segments.

6. The dynamic waveform blood pressure calibration system for an invasive blood pressure monitor according to claim 1, characterized in that, The fingerprint registration module includes a frequency extraction submodule, a peak comparison submodule, and an interval mapping submodule. The frequency extraction submodule is used to call the resonance candidate fragment set, extract feature parameters such as duct resonance frequency, oscillation start and stop pressure, and oscillation peak amplitude, statistically analyze the amplitude decay law, and generate an oscillation frequency sequence. The peak comparison submodule is used to compare the time interval and amplitude changes of adjacent oscillation peaks based on the oscillation frequency sequence, determine the oscillation continuity state, and obtain the peak association interval. The interval mapping submodule is used to match the peak association interval with the pressure interval of the pressure range calibration table, mark the resonance feature attributes, and classify and establish a graded resonance fingerprint table.

7. The dynamic waveform blood pressure calibration system for an invasive blood pressure monitor according to claim 1, characterized in that, The waveform verification module includes a frequency call submodule, a waveform pressure comparison submodule, and a waveform calibration submodule. The frequency call submodule is used to call the graded resonance fingerprint table, load the frequency, actuation duration and output pressure of the piezoelectric ceramic high-frequency actuator and the dual-stage gas-liquid isolation chamber, analyze the waveform change state and generate the actuation waveform sequence; The waveform pressure comparison submodule is used to compare the actuation waveform sequence with the standard blood pressure waveform sampling data, calculate the pressure deviation, screen the waveform range with the matching degree meets the standard, and obtain the waveform matching interval. The waveform calibration submodule is used to align waveform timing, amplitude, and period parameters based on waveform matching intervals, configure waveform feature labels, and establish standard input waveform identifiers.

8. The dynamic waveform blood pressure calibration system for an invasive blood pressure monitor according to claim 1, characterized in that, The compensation database module includes an intensity acquisition submodule, a phase comparison submodule, and a compensation table creation submodule. The intensity acquisition submodule is used to acquire the standard input intensity value and the detected output intensity value based on the standard input waveform identifier, calculate the amplitude deviation, and generate an intensity offset sequence. The phase comparison submodule is used to extract the standard phase value and the detected output phase value, calculate the phase time difference, analyze the correlation characteristics between phase and intensity, and obtain the phase offset interval. The compensation table creation submodule is used to integrate parameters such as amplitude correction, timing compensation, and frequency response correction, and combine them with pressure levels to complete parameter matching and create a dynamic calibration compensation table.

9. The system according to claim 1, characterized in that, The process of screening the sampling segment after the steep rise edge and the sampling segment adjacent to the notch also includes: when the time interval between adjacent pressure peaks meets the preset stability condition and the amplitude change meets the preset attenuation condition, the corresponding sampling interval is marked as a resonance candidate segment.

10. A method for calibrating dynamic waveform blood pressure in an invasive blood pressure monitor, applied to the system described in any one of claims 1 to 9, characterized in that, The method includes the following steps: Pressure profile establishment steps: Obtain the stroke value of the micro motor plunger pump, the opening value of the proportional valve, and the output pressure value of the two-stage gas-liquid isolation chamber; compare the basic pressure set value with the output pressure value; and generate a pressure profile calibration table. Segment identification steps: Call the pressure range calibration table to collect standard input waveform pressure values, systolic peak pressure, and dicrotic notch pressure; filter the sampling segments after the steep rise edge and the sampling segments adjacent to the notch to generate a set of resonance candidate segments; Fingerprint registration steps: Call the set of resonance candidate segments, obtain the duct resonance frequency, oscillation start and stop pressure, oscillation duration, and oscillation peak amplitude, compare the time interval and amplitude changes of adjacent oscillation peaks, and generate a graded resonance fingerprint table; Waveform verification steps: Call the frequency value of the graded resonance fingerprint table and the piezoelectric ceramic high-frequency actuator, compare the output pressure value of the two-stage gas-liquid isolation chamber with the blood pressure waveform pressure sampling value, and generate a standard input waveform identifier; Compensation entry steps: Call the standard input waveform identifier to obtain the output pressure value of the invasive blood pressure monitor, the standard input intensity value, the tested output intensity value, and the input phase value. Calculate the intensity difference and phase time difference to establish a dynamic calibration compensation table.

Images