Method and system for assessing bleeding risk in general surgery resection

Abstract

The invention relates to the technical field of operation risk assessment, in particular to a bleeding risk assessment method and system in general surgery resection, and the method comprises the following steps: collecting the instantaneous pressure of the outer wall of a blood vessel in the operation to extract the main frequency and energy characteristic calibration time, collecting the multi-direction strain rate to synthesize a fracture expansion rate curve, and extracting the peak value and intensity; the method comprises the following steps: acquiring a pressure value, calculating a key time point phase difference, combining with intensity to generate a fracture critical state, matching a frequency band and performing frequency conversion to judge a resonance generation response section, weighting and calculating a risk index in combination with the fracture state, and outputting a result when a threshold value is exceeded and a bleeding risk label is added. Frequency domain energy and strain rate are extracted to quantify a fracture trend, a phase difference is analyzed to judge a fracture state, resonance is identified by combining frequency conversion, a risk index is calculated by fusing multi-source signal weighting, accurate risk prediction and dynamic response are realized, and support is provided for operation decision making.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of surgical risk assessment, in particular to a general surgery resection bleeding risk assessment method and system. BACKGROUND

[0002] The technical field of surgical risk assessment includes the analysis, prediction and evaluation of various risks during surgery, with a particular focus on the effective evaluation of bleeding risks, organ damage, anesthetic reactions and other risks during surgical operations. The core content includes precise prediction of surgical risks through various technical means, such as medical imaging, vital sign monitoring, patient history analysis, etc., and provides a scientific basis for the selection of surgical plans. The purpose is to reduce unforeseen risks during surgery through reasonable evaluation methods, ensure patient safety, and improve the success rate of surgery.

[0003] Among them, the general surgery resection bleeding risk assessment method and system refers to a technical solution for evaluating the bleeding risk that may occur during general surgery resection. The technical matters cover quantitative analysis of bleeding risk based on patient physical examination data, preoperative imaging examination results, medical history information and other data. Specifically, a bleeding risk prediction model is constructed, and the bleeding probability is evaluated using statistical analysis methods, machine learning algorithms and other means in combination with patient individual characteristics, surgical types, surgical sites and other multi-dimensional factors.

[0004] Existing technologies rely on preoperative images, medical history information and static vital sign parameters to quantitatively analyze bleeding risks. Although it has certain value in forming a preliminary judgment of risks in the early stage of surgery, it lacks dynamic capture of microstructure changes and real-time physiological parameters, and cannot respond to rapidly evolving risks. This analysis method based on pre-set models and patient historical data has significant lag when facing intraoperative emergencies, and cannot perceive real-time physical changes such as crack propagation and structure rupture. For example, when the blood vessel wall is critically strained due to operation traction, the existing method cannot immediately identify the change and respond, which easily causes sudden bleeding without warning. On the other hand, the risk results in current bleeding risk assessment are mostly divided by static labels, without establishing a continuous risk evolution mechanism based on real-time parameter changes, resulting in a coarse risk assessment granularity that cannot be refined to specific operation nodes or specific time periods. The above shortcomings are particularly evident in complex general surgery, and the lag in risk control not only affects the doctor's response strategy, but also poses a potential threat to patient safety. SUMMARY

[0005] In order to solve the technical problems existing in the prior art, the general surgery resection bleeding risk assessment method provided by the embodiments of the present application includes the following steps: In order to achieve the above object, the present application adopts the following technical scheme: the general surgery resection intraoperative bleeding risk assessment method comprises the following steps: S1: collecting the instantaneous pressure value of the blood vessel outer wall in the operation area during resection and calculating the difference, extracting the main frequency and frequency energy value of the difference result and performing time sequence calibration to generate the pressure main frequency feature; S2: based on the time label corresponding to the pressure main frequency feature, collecting the multi-directional strain rate signal of the blood vessel outer wall and performing vector synthesis as the crack propagation rate curve, extracting the propagation rate peak time point and calculating the intensity level to generate the crack propagation rate peak feature; S3: obtaining the corresponding key time points of the crack propagation rate peak feature and the pressure main frequency feature and calculating the phase difference, jointly analyzing the phase difference and the crack propagation rate intensity level to generate the fracture critical state; S4: based on the fracture critical state matching the energy concentration frequency band, combining the corresponding crack propagation rate curve and performing frequency conversion, and combining the energy concentration frequency band to judge whether there is resonance, and generating the resonance response section; S5: based on the resonance response section and the fracture critical state, the risk index is calculated by weighting, if the risk index exceeds the risk level threshold interval, the bleeding risk label is added, and the general surgery resection intraoperative bleeding risk result is generated.

[0006] As a further scheme of the present application, the pressure main frequency feature includes the main frequency value, the frequency energy distribution, and the time calibration information, the crack propagation rate peak feature includes the propagation rate peak value and the intensity level, the fracture critical state includes the key time point, the phase difference value, and the fracture strength threshold, the resonance response section includes the frequency conversion result, the energy concentration frequency band, and the resonance identification interval, and the bleeding risk result includes the risk index value, the risk level threshold, and the bleeding risk label.

[0007] As a further scheme of the present application, the specific steps of S1 are: S101: collecting the instantaneous pressure value of the blood vessel outer wall in the operation area during resection, and calculating the difference of the pressure values of adjacent sampling points, and reorganizing according to the sampling time sequence to obtain the differential pressure sequence; S102: based on the differential pressure sequence, performing fast Fourier transform operation, extracting the amplitude peak point in the frequency spectrum distribution and determining the corresponding frequency value, and simultaneously performing cumulative operation on the energy of multiple frequency components to obtain the frequency energy parameter set; S103: calling the frequency energy parameter set, corresponding time sequence calibration of the main frequency value and the frequency energy vector according to the time sampling index, and collecting the calibrated frequency value as a characteristic factor to generate the pressure main frequency feature.

[0008] As a further scheme of the present application, the specific step of S2 is: S201: Based on the time mark corresponding to the pressure main frequency feature, collect the multi-directional strain rate signals of the extravascular wall under the same time index, decompose the multi-directional signals in the form of vectors according to the coordinate components, and then synthesize the corresponding components to obtain a synthesized strain rate curve; S202: Call the synthesized strain rate curve, retrieve the rate values of the multiple time sampling points in the curve, extract the extended rate peak time points, calculate the difference intensity of the preset extended rate reference value, and obtain a peak time intensity group; S203: According to the peak time intensity group, bind the time point index with the intensity difference, and use the bound time point and intensity value as the feature output to generate the crack extension rate peak feature.

[0009] As a further scheme of the present application, the extended rate reference value is set by arithmetically averaging all the rate values in the sampling interval before and after the time point corresponding to the maximum rate value.

[0010] As a further scheme of the present application, the specific step of S3 is: S301: Obtain the time index of the crack extension rate peak feature and the pressure main frequency feature, retrieve the key sampling points corresponding to the time indexes of the two features, normalize the time difference of the corresponding points to obtain the phase angle, and calculate the phase difference sequence in time sequence; S302: Call the phase difference sequence and pair the intensity levels recorded in the crack extension rate peak feature point by point, perform joint distribution calculation on multiple pairs of paired data to obtain a phase intensity joint distribution; S303: According to the phase intensity joint distribution, analyze the phase intensity distribution density of multiple points and combine the phase density threshold to screen the critical interval, aggregate the phase difference and intensity value corresponding to the critical interval, and generate a fracture critical state.

[0011] As a further scheme of the present application, the phase density threshold is set by statistically analyzing the density of the joint distribution results formed after the phase difference and intensity level are paired, retrieving the distribution frequency of multiple value points and calculating the frequency average value, and combining the multiple coefficient of the reference value.

[0012] As a further scheme of the present application, the specific step of S4 is: S401: Based on the concentrated frequency section recorded in the fracture critical state matching energy spectrum data, extract the spectrum energy value of the section and perform continuous interval aggregation to obtain an energy concentrated frequency section; S402: Call the energy concentration frequency band, and combine the time domain sampling points of the crack propagation rate peak signal to perform Fourier transform on the signal to obtain frequency components, and then map the frequency components to the energy concentration frequency band to generate a frequency conversion result; S403: Perform resonance determination according to the frequency conversion result and the numerical interval of the energy concentration frequency band, and if the frequency components form a resonance relationship with the energy concentration frequency band, record the corresponding time period as a response section to generate a resonance response section.

[0013] As a further scheme of the present application, the specific steps of S5 are: S501: Based on the corresponding time index of the resonance response section and the fracture critical state, perform weighted product operation on the spectral amplitude values recorded in the resonance response section and the stress intensity recorded in the fracture critical state, and perform addition normalization on the product result to obtain a risk index sequence; S502: Call the risk index sequence, and perform interval comparison on the risk index values of multiple time points and the preset risk level threshold interval, mark the time point positions of all risk index values exceeding the risk level threshold interval, and generate a risk label index set; S503: According to the corresponding relationship of the risk label index set and the risk index sequence, integrate the risk index values of the marked positions into a bleeding risk record data frame, and attach a bleeding risk label to generate a general surgery resection intraoperative bleeding risk result; The risk level threshold interval is set by hierarchical statistics on the numerical distribution of the risk data frame during the operation.

[0014] The general surgery resection intraoperative bleeding risk assessment system comprises: A pressure analysis module acquires the instantaneous pressure value of the blood vessel outer wall in the operation area during resection and calculates the difference, extracts the main frequency and frequency energy value of the difference result, and performs time sequence calibration to generate a pressure main frequency feature and transmit it to a crack feature module; The crack feature module acquires the multi-directional strain rate signal of the blood vessel outer wall based on the corresponding time label of the pressure main frequency feature, and performs vector synthesis as a crack propagation rate curve, extracts the propagation rate peak time point and calculates the intensity level, generates a crack propagation rate peak feature, and transmits it to a critical judgment module; The critical judgment module obtains the corresponding key time points of the crack propagation rate peak feature and the pressure main frequency feature and calculates the phase difference, performs joint analysis on the phase difference and the crack propagation rate intensity level, generates a fracture critical state, and transmits it to a response recognition module; A response identification module matches an energy concentration frequency band based on the fracture critical state, combines a corresponding crack propagation rate curve and performs frequency conversion, judges whether resonance occurs in combination with the energy concentration frequency band, generates a resonance response section and delivers it to a risk assessment module; A risk assessment module performs weighted calculation of a risk index based on the resonance response section and the fracture critical state, and if the risk index exceeds a risk level threshold interval, an additional bleeding risk label is added, and a general surgery resection intraoperative bleeding risk result is generated.

[0015] Compared with the prior art, the advantages and positive effects of the present application are that: In the present application, by collecting the instantaneous pressure value of the blood vessel outer wall in the operation area during resection and calculating the difference, the dynamic stress state of the blood vessel wall can be tracked in real time, and the process of extracting the main frequency and frequency energy value makes it highly sensitive to local microscopic changes, so that potential bleeding signals can be captured at an early stage. The multi-directional strain rate signal extracted based on the time marker further realizes the multi-dimensional vector synthesis of the crack propagation rate, and through the extraction of the peak point and intensity level, the dynamic trend of structure rupture is quantitatively expressed. In this way, the key rupture behavior is no longer dependent on static images or traditional indicators, but on the risk evolution trajectory captured based on continuously changing physical signals. Through the coupling analysis of phase difference and intensity level, the key energy release stage in the crack propagation process can be identified, so as to judge the fracture critical state, and then whether structure resonance occurs is judged in combination with frequency conversion and energy concentration frequency band analysis. A dynamic response mechanism at the structure mechanics level is introduced in the bleeding risk judgment. Finally, by means of weighted risk index evaluation of the fracture state and the resonance section, a quantitative label highly related to the actual risk of bleeding is established, so that the evaluation result not only has timeliness and dynamics, but also has precise prediction ability. In this overall processing logic, through real-time physical parameter acquisition, collaborative processing of frequency domain and time domain signals, interactive analysis between multiple sources of signals, and risk quantitative evaluation based on dynamic response, the problem of insufficient risk perception of sudden and dynamic risks in the past evaluation methods is effectively solved, the accuracy and reaction speed of risk identification are improved, and highly reliable data support is provided for clinical operation decision-making. BRIEF DESCRIPTION OF DRAWINGS

[0016] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed in the embodiment description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0017] Figure 1 The step flowchart of the present application; Figure 2S1 refinement schematic diagram of the present application; Figure 3 S2 refinement schematic diagram of the present application; Figure 4 S3 refinement schematic diagram of the present application; Figure 5 S4 refinement schematic diagram of the present application; Figure 6 S5 refinement schematic diagram of the present application; Figure 7 System module diagram of the present application. DETAILED DESCRIPTION

[0018] The technical solutions in the present application will be described below with reference to the drawings.

[0019] In the embodiments of the present application, the words such as "example", "for example" are used to represent as an example, illustration or description. Any embodiment or design scheme described as "example" in the present application should not be interpreted as more preferred or more advantageous than other embodiments or design schemes. Rather, the word "example" is intended to present the concept in a specific manner. In addition, in the embodiments of the present application, the meaning expressed by "and / or" can be both, or can be one of the two.

[0020] In the embodiments of the present application, "image" and "picture" can be used interchangeably at times. It should be pointed out that when the distinction is not emphasized, the meanings expressed are consistent. "Of", "corresponding" and "corresponding" can be used interchangeably at times. It should be pointed out that when the distinction is not emphasized, the meanings expressed are consistent.

[0021] In the embodiments of the present application, sometimes the subscript such as W1 can be written in the form of non-subscript such as W1. When the distinction is not emphasized, the meanings expressed are consistent.

[0022] In order to make the technical problems, technical solutions and advantages to be solved by the present application more clear, the following will be described in detail with reference to the drawings and specific embodiments.

[0023] Please refer to Figure 1 The embodiment of the present application provides a general surgery resection bleeding risk assessment method, which comprises the following steps: S1: Collecting the instantaneous pressure value of the blood vessel outer wall in the operation area during resection and calculating the difference, extracting the main frequency and frequency energy value of the difference result and performing time sequence calibration to generate a pressure main frequency feature; S2: Based on the time mark corresponding to the pressure main frequency feature, collect the multi-directional strain rate signal of the blood vessel wall and perform vector synthesis as the crack propagation rate curve, extract the peak time point of the propagation rate and calculate the intensity level, and generate the crack propagation rate peak feature; S3: Obtain the corresponding key time point of the crack propagation rate peak feature and the pressure main frequency feature and calculate the phase difference, jointly analyze the phase difference and the crack propagation rate intensity level, and generate the fracture critical state; S4: Based on the fracture critical state matching energy concentration frequency band, combine the corresponding crack propagation rate curve and perform frequency conversion, and combine the energy concentration frequency band to determine whether it is resonant, and generate the resonance response section; S5: Based on the resonance response section and the fracture critical state, the risk index is calculated by weighting, and if the risk index exceeds the risk level threshold interval, the bleeding risk label is added, and the bleeding risk result in the general surgery resection is generated.

[0024] The pressure main frequency feature includes the main frequency value, the frequency energy distribution, and the time calibration information, the crack propagation rate peak feature includes the propagation rate peak value and the intensity level, the fracture critical state includes the key time point, the phase difference value, and the fracture strength threshold, the resonance response section includes the frequency conversion result, the energy concentration frequency band, and the resonance recognition interval, and the bleeding risk result includes the risk index value, the risk level threshold, and the bleeding risk label.

[0025] Please refer to Figure 2 , the specific steps of S1 are: S101: Collect the instantaneous pressure value of the blood vessel wall in the operation area of the resection, and perform difference calculation on the pressure values of adjacent sampling points, and reorganize according to the sampling time sequence to obtain the difference pressure sequence; In the operation area of the resection, a section of the target blood vessel wall that is not blocked and easy to access is selected, and three piezoelectric sensor units are arranged along the blood vessel axis at equal intervals of 5 mm, denoted as P1, P2, and P3. During the operation of the surgical instrument, the instantaneous pressure values of each unit are synchronously collected at a frequency of 1000 Hz for 5 seconds. For example, at a certain time, the collected part of the original pressure data is as follows: at time stamp 101 milliseconds, the pressure values of P1, P2, and P3 are 13.25, 13.88, and 13.54 kilopascals, respectively; at time stamp 102 milliseconds, the pressure values are 13.28, 13.90, and 13.56 kilopascals, respectively; and at time stamp 103 milliseconds, the pressure values are 13.31, 13.91, and 13.59 kilopascals, respectively. After the collection is completed, the difference between adjacent sampling points in the pressure sequence obtained by each sensor is calculated. The difference between the pressure value of the current sampling point and the pressure value of the previous sampling point is calculated. Taking sensor P1 as an example, the differential pressure value at time stamp 102 milliseconds is obtained by subtracting the pressure value of 13.25 kilopascals at time stamp 101 milliseconds from the pressure value of 13.28 kilopascals at time stamp 102 milliseconds, and the result is 0.03 kilopascals. Similarly, the differential pressure value at 103 milliseconds is 13.31-13.28=0.03 kilopascals. This difference calculation is performed for all sampling points of all sensors to form three independent differential pressure sequences. Finally, the three sequences are recombined in the order of sampling time, that is, at each time stamp, the differential pressure values of the three sensors are combined into a data group, thereby obtaining a differential pressure sequence reflecting the pressure change rate of each monitoring point.

[0026] S102: Perform a fast Fourier transform operation based on the differential pressure sequence, extract the amplitude peak points in the frequency spectrum distribution and determine the corresponding frequency values, and simultaneously perform an energy accumulation operation on the multi-frequency components to obtain a frequency energy parameter set; For each dimension in the generated differential pressure sequence, 4096 consecutive sampling points are intercepted as an analysis window, and a fast Fourier transform is independently performed to convert the time domain signal into an amplitude spectrum in the frequency domain. To accurately extract the peak point from the amplitude spectrum, a predetermined amplitude threshold is set. The setting process of the threshold is as follows: 20 cases of resting state blood vessel pressure signals without surgical instrument interaction are collected, and after the same processing, the average noise amplitude of all signals is calculated to be 0.08 units (AU) in the frequency range of 80-200 Hz, and the average standard deviation is 0.02 units. The threshold is set to be the average value plus three times the standard deviation, that is, 0.08 plus 3 times 0.02, which is equal to 0.14 AU. When analyzing the amplitude spectrum, the amplitude of a frequency point must be greater than the amplitudes of the two adjacent frequency points, and the amplitude itself must be greater than 0.14 AU, in order to be determined as a peak point. For example, if the amplitude at 1.5 Hz is 0.85 AU, and the adjacent point amplitudes are 0.21 AU and 0.18 AU, then 1.5 Hz is determined as a peak point because 0.85 is greater than the adjacent values and greater than 0.14 AU. Subsequently, the identified peak points are accumulated in frequency bands, and the frequency bands are divided into: physiological frequency band (0.5-5 Hz), instrument operation frequency band (5-50 Hz), and high-frequency noise frequency band (50-250 Hz). The energy is the sum of the squares of the amplitudes of the peaks in the frequency band. A specific example is given as follows: two peak points, 1.5 Hz (amplitude 0.85) and 2.5 Hz (amplitude 0.17), are identified in the physiological frequency band, and the energy of the frequency band is the sum of the squares of 0.85 and 0.17, i.e. 0.7225 plus 0.0289, equal to 0.7514; two peak points, 18.5 Hz (amplitude 0.32) and 30.1 Hz (amplitude 0.25), are identified in the instrument operation frequency band, and the energy is the sum of the squares of 0.32 and 0.25, i.e. 0.1024 plus 0.0625, equal to 0.1649; two peak points, 60.2 Hz (amplitude 0.11) and 120.0 Hz (amplitude 0.09), are identified in the high-frequency noise frequency band, and the energy is the sum of the squares of 0.11 and 0.09, i.e. 0.0121 plus 0.0081, equal to 0.0202. Finally, the frequency corresponding to the highest amplitude point among all amplitude peak points (0.85 AU at 1.5 Hz in this example) is taken as the main frequency value, i.e. 1.5 Hz, and the energy accumulation values of the frequency bands [0.7514, 0.1649, 0.0202] form a frequency energy vector, which together constitutes the frequency energy parameter set of the data window.

[0027] S103: Call the frequency energy parameter set, and time sequence mark the main frequency value and the frequency energy vector according to the time sampling index, and collect the marked frequency value as a characteristic factor to generate a pressure main frequency characteristic; The generated frequency energy parameter set is called to realize time sequencing in a sliding window manner. A complete differential pressure sequence is analyzed in segments by using an analysis window with a length of 1024 sampling points and a sliding step of 256 sampling points. For each window, the operation of S102 is repeated to generate a set of frequency energy parameters. Then, each set of parameters is corresponded with the center point timestamp of the window to complete time sequencing. For example, the first window covers 0 to 1.024 seconds, and the center timestamp is 0.512 seconds; the second window covers 0.256 to 1.280 seconds, and the center timestamp is 0.768 seconds. The main frequency values obtained by analysis at the center timestamps 0.512 seconds and 0.768 seconds are both 1.5 Hz, and the energy of the instrument operation frequency band is lower than 0.2. At the center timestamp 1.280 seconds, the main frequency value suddenly changes to 18.2 Hz, and the energy of the instrument operation frequency band jumps to 2.88, which indicates that the surgical instrument interacts with the blood vessel. To define the physical meaning of the main frequency value, experiments are conducted for verification. In 10 simulated surgeries, the main frequency values at the baseline state are distributed in 1.2 to 1.6 Hz, and the main frequency values during instrument operation are distributed in 18.5 to 20.1 Hz. Based on this, 5 Hz is taken as the limit to distinguish physiological activity from instrument interaction. Then, the main frequency value of each timestamp after labeling is taken as a characteristic factor to be collected, forming a time sequence, such as (0.512 seconds, 1.5 Hz), (0.768 seconds, 1.5 Hz), (1.280 seconds, 18.2 Hz), etc. The advantage of this way is that by performing time-frequency analysis on the differential pressure sequence and extracting the main frequency value as a feature, weak pressure changes can be converted into significant frequency features, so as to accurately track the dynamic interaction process between the surgical instrument and the blood vessel wall. Finally, the main frequency time sequences collected by all sensors are integrated to generate a pressure main frequency feature.

[0028] Please refer to Figure 3 The specific steps of S2 are as follows: S201: Based on the time label corresponding to the pressure main frequency feature, the multi-directional strain rate signals of the blood vessel wall at the same time index are collected, the multi-directional signals are decomposed into coordinate components in vector form, and the corresponding components are vector synthesized to obtain a synthesized strain rate curve; Based on the generated pressure dominant frequency feature, time points with dominant frequency values higher than 5 Hz are selected, which are marked as instrument interaction event points. Take a time marker in the pressure dominant frequency feature, for example, 1.280 seconds (its dominant frequency value is 18.2 Hz), as a reference, collect the extravascular wall multi-directional strain rate signals synchronized with this time index. The signals are collected by a 0-45-90 degree rectangular rosette strain gauge containing three independent strain gauges deployed co-located with the pressure sensor array. The three strain gauges measure the strain rate of the extravascular wall along the axial direction (0 degree direction), 45 degree direction, and circumferential direction (90 degree direction), respectively, denoted as axial strain rate, 45 degree strain rate, and circumferential strain rate. These signals are recorded at the same 1000 Hz frequency as the pressure signal, with units of per second (s ). Then, the collected multi-directional strain rate signals are decomposed into coordinate components in vector form. A two-dimensional Cartesian coordinate system with the monitoring point as the origin is established, where the x-axis is along the axial direction of the blood vessel, and the y-axis is along the circumferential direction of the blood vessel. In this coordinate system, the axial strain rate is directly taken as the strain rate component in the x-direction, and the circumferential strain rate is directly taken as the strain rate component in the y-direction. Then, the corresponding components are vector synthesized. The synthesis strain rate is calculated by taking the square root of the sum of the squares of the x-direction and y-direction strain rate components. This calculation is performed for each time sampling point, generating a synthesis strain rate curve that changes over time. Near the time stamp 1.280 seconds, the collected part of the multi-directional strain rate signal data is shown in the following table.

[0029] Table 1: Multi-directional strain rate time series data table Time stamp (seconds) Axial strain rate (1 / s) ) Circumferential strain rate (εθ) ) 1.279 0.75 1.42 1.280 0.80 1.50 1.281 0.95 1.88 1.282 1.12 2.21 1.283 1.20 2.35 1.284 1.15 2.26 1.285 1.08 2.12 As shown in Table 1, part of the strain rate data near the key time points is listed. Taking the data at time stamp 1.280 seconds as an example, the axial strain rate is 0.80 s , and the circumferential strain rate is 1.50 s . The synthesis strain rate is calculated by taking the square root of the sum of the squares of 0.80 and 1.50, resulting in 1.70 . At time stamp 1.283 seconds, the axial strain rate is 1.20 , and the circumferential strain rate is 2.35 , and the synthesis strain rate is calculated by taking the square root of the sum of the squares of 1.20 and 2.35, resulting in approximately 2.64 . Repeat this calculation for all time sampling points to ultimately form a synthesis strain rate curve.

[0030] S202: Call the synthesis strain rate curve, retrieve the rate values of multiple time sampling points in the curve, and extract the extended rate peak time points, calculate the difference strength with the preset extended rate reference value, and obtain the peak time strength group; The generated synthetic strain rate curve is called upon to perform data retrieval to identify and quantify the peak events of the rate. First, the curve is subjected to peak point extraction. A time sampling point is determined as an extended rate peak point if it satisfies two conditions: first, the rate value at this point is greater than all the rate values within a 5-sample neighborhood window (i.e., a 10-millisecond neighborhood window); second, the rate value at this point must be higher than a preset peak identification threshold. The peak identification threshold is set to 0.20 , which is set based on statistical analysis of the synthetic strain rate data of 50 cases of non-instrument interaction phases in surgery, finding that 99.8% of the peak values of physiological fluctuations are lower than this value. The setting process of this threshold is as follows: 50 sets of baseline strain data with a duration of 10 seconds are collected, the synthetic strain rate curve is calculated, all local peaks are extracted, forming a set containing 857 peaks, the mean value of the set is 0.06 , and the standard deviation is 0.045 . The threshold value is taken as the mean value plus about 3.1 times the standard deviation, 0.06 + 3.1 x 0.045 ≈ 0.20 . Then, the difference intensity of the extracted extended rate peak value and a preset extended rate reference value is calculated. The extended rate reference value here represents the upper limit of the normal strain rate that the tissue can produce by itself under the physiological activity (such as heart beat) without external instrument action. The setting process of the reference value is as follows: in 25 cases of isolated perfused porcine aorta experiments, the physiological conditions of normal blood pressure (120 / 80 mmHg) and heart rate (75 times / minute) are simulated, and the strain rate signal of the blood vessel wall is continuously collected. The synthetic strain rate curve is calculated by the method described in S201, and the maximum synthetic strain rate in each cardiac cycle is recorded. The maximum rate values of all cardiac cycles (about 1800 data points in total) are statistically analyzed, and the 99th percentile is taken as the extended rate reference value. Experimental data shows that the value is 0.12 . Specific example: refer to the synthetic strain rate curve generated in S201, at time point 1.283 seconds, the rate value is 2.64 . Check the 5 sampling points before and after, find that the rate value is the maximum value in the neighborhood, and 2.64 is much larger than the peak identification threshold 0.20 , therefore, 1.283 seconds is determined as an extended rate peak time point, and the peak value is 2.64 . Then, the difference intensity of the peak value and the extended rate reference value is calculated: extended rate peak value: 2.64 . Extended rate reference value: 0.12 . Difference intensity = 2.64 - 0.12 = 2.52 s If another peak point satisfying the condition is identified in the subsequent part of the curve, for example, at time point 1.950 seconds, with a rate value of 1.85 , then the difference intensity is also calculated: difference intensity = 1.85 s - 0.12 s = 1.73 . Through the search and calculation of the entire synthetic strain rate curve, a peak time intensity group composed of multiple peak time points and corresponding difference intensities is obtained. For example, the obtained result group is (1.283 seconds, 2.52 s ), (1.950 seconds, 1.73 s ). The advantage of this method is that by calculating the difference between the expansion rate peak value and the physiological reference value, the impact strength of the surgical instrument operation on the blood vessel wall can be quantified, excluding the interference of physiological strain fluctuations. The results show that a rapid deformation event with an intensity of 2.52 s occurred at 1.283 seconds.

[0031] S203: According to the peak time intensity group, the time point index is bound with the intensity difference, and the bound time point and intensity value are taken as the feature output to generate the crack expansion rate peak feature; According to the generated peak time intensity group, the data in it is structured and bound. The specific execution process is that each identified peak event is taken as an independent data unit, which contains two bound data items inside: the time point index of the peak occurrence, and the difference intensity corresponding to the peak. This binding operation integrates the originally dispersed time and intensity information into a feature pair with clear physical meaning. Taking the calculation result of S202 as an example, the peak time intensity group is [(1.283 seconds, 2.52 ), (1.950 seconds, 1.73 )]. After the binding operation, two feature units are formed: the first feature unit: the time point index is 1.283 seconds, and the intensity difference is 2.52 . The second feature unit: the time point index is 1.950 seconds, and the intensity difference is 1.73 . All the time points and intensity values after binding are output as features. The result of this output is a structured list or array, where each element represents a time and quantified intensity of a significant blood vessel wall rapid deformation event. This list is the final generated crack propagation rate peak feature. Specific example: based on the results of S202, the generated crack propagation rate peak feature is: crack propagation rate peak feature = (1.283, 2.52), (1.950, 1.73) where the first element of each tuple is the time (unit: seconds), and the second element is the difference intensity (unit: ). The generation of this feature does not rely on continuous strain rate data streams, but rather extracts the complex dynamic process into a few key, high-information-content discrete event points. The intensity difference 2.52 represents a high-speed deformation far beyond the normal physiological range, and the numerical size is directly related to the severity of tissue tearing or injury. A higher intensity difference, for example, greater than 2.0 , is determined as a "high-risk" event; a value between 1.0 and 2.0 is determined as a "moderate-risk" event; and a value less than 1.0 is a "low-risk" event. The division of risk intervals is based on the retrospective analysis of 100 intraoperative data of different degrees of blood vessel tearing, and the distribution range of difference intensity corresponding to different damage levels is counted. The advantage of this method is that by converting the dynamic strain rate curve into a set of discrete feature points containing time and intensity, the data volume is greatly compressed, and the key spatiotemporal information of potential tissue damage events is directly revealed. The output crack propagation rate peak feature provides direct and quantitative input for subsequent risk assessment and early warning.

[0032] Please refer to Figure 4 , the specific steps of S3 are: S301: Obtain the time index of the crack propagation rate peak feature and the pressure main frequency feature, retrieve the key sampling points corresponding to the time index of the two, normalize the time difference of the corresponding points to obtain the phase angle, and calculate the phase difference sequence in chronological order; First, obtain the generated crack propagation rate peak feature, that is, the data pairs containing time points and intensity differences, such as (1.283 seconds, 2.52 ) and (1.950 seconds, 1.73 ). At the same time, the pressure dominant frequency time series generated in S103 is obtained, which contains data points like (1.280 seconds, 18.2 Hz). Then, based on each time point of the crack propagation rate peak, the pressure dominant frequency event point closest in time and earlier than the peak point is found by searching forward in the pressure dominant frequency time series. For example, for the rate peak of 1.283 seconds, the corresponding pressure event point is 1.280 seconds, and the dominant frequency is 18.2 Hz. Next, the difference between the two time points is calculated, i.e. 1.283 seconds minus 1.280 seconds, which is 0.003 seconds. Then, the time difference is normalized to obtain the phase angle. The period for conversion is based on the dominant frequency of the corresponding pressure event, and the period is the inverse of the frequency, i.e. 1 divided by 18.2 Hz, which is approximately 0.0549 seconds. The calculation process of the phase angle is to divide the time difference 0.003 seconds by the period 0.0549 seconds, and then multiply the quotient by 360 degrees to obtain approximately 19.7 degrees. Repeat this process for the second rate peak event (1.950 seconds), set its corresponding pressure event point to 1.945 seconds, and the dominant frequency to 19.5 Hz. The time difference is 0.005 seconds, and the dominant frequency period is approximately 0.0513 seconds. The calculated phase angle is approximately 35.1 degrees. Finally, arrange all the calculated phase angles in chronological order to obtain the phase difference sequence [19.7 degrees, 35.1 degrees].

[0033] S302: Call the phase difference sequence and pair each point with the intensity level recorded in the crack propagation rate peak feature, perform joint distribution calculation on multiple pairs of paired data to obtain the phase intensity joint distribution; Call the generated phase difference sequence [19.7 degrees, 35.1 degrees] and pair each point with the intensity level [2.52 , 1.73 ] recorded in the crack propagation rate peak feature. The first event pair is (phase difference 19.7 degrees, intensity level 2.52 ), and the second event pair is (phase difference 35.1 degrees, intensity level 1.73 ). To perform joint distribution calculation, set three more pairs of paired data obtained in subsequent monitoring: (25.5 degrees, 2.81 ), (18.9 degrees, 2.65 ) and (95.2 degrees, 0.88 ). Then, perform joint distribution calculation on the five pairs of paired data. This calculation is done by constructing a two-dimensional grid, with the phase difference dimension of the grid ranging from 0 degrees to 180 degrees, divided into an interval every 20 degrees; the intensity level dimension ranges from 0 to 3.0 , every 0.5 is divided into an interval. Then, each data pair is assigned to its corresponding grid cell, and the number of data points in each cell is counted. For example, the data pair (19.7 degrees, 2.52 ) falls into the cell of phase interval [0 degree, 20 degree] and intensity interval [2.5 , 3.0 ]. After the classification and counting of all five data points, the number of data points in each cell is obtained. The cell of phase interval [0 degree, 20 degree] and intensity interval [2.5 , 3.0 ] contains 2 data points; the cell of phase interval [20 degree, 40 degree] and intensity interval [1.5 , 2.0 ] contains 1 data point; and the cell of the same phase interval but intensity interval [2.5 , 3.0 ] also contains 1 data point. This set of grid cells and their corresponding data point count values constitute the phase-intensity joint distribution.

[0034] S303: According to the phase-intensity joint distribution, analyze the phase-intensity distribution density of the multi-point, combine with the phase density threshold, and screen the critical interval, aggregate the phase difference and intensity value corresponding to the critical interval, and generate the fracture critical state; According to the generated phase-intensity joint distribution, analyze the distribution density of each cell, i.e., the number of data points. To screen the critical interval with a high concentration of data points, a phase density threshold is set. The setting of this threshold is determined by retrospective analysis of 30 cases of surgical data that have occurred blood vessel tearing. In these cases, the mean value of the count value of the highest density cell in the phase-intensity joint distribution within the 5-second window before tearing is 2.8, and the standard deviation is 0.6. The phase density threshold is set to the mean value minus one standard deviation, i.e., 2.2, and the integer 3. Therefore, any cell with a density value of 3 or more is determined to be a critical interval. Returning to the calculation of this embodiment, in addition to the five data points set in S302, a new data point (17.5 degrees, 2.8 ) is obtained, which also falls into the cell of phase interval [0 degree, 20 degree] and intensity interval [2.5 , 3.0 ], making the density value of this cell change from 2 to 3. Since this density value 3 is equal to the set threshold 3, this cell is screened as a critical interval. Then, all three data points falling into this critical interval, i.e., (19.7 degrees, 2.52 ), (18.9 degrees, 2.65 ), and (17.5 degrees, 2.8 ), the aggregate is calculated. The aggregate is the arithmetic average of the phase difference and the intensity value. The aggregate phase difference is the sum of the three divided by three, which is 18.7 degrees. The aggregate intensity value is the sum of the three divided by three, which is approximately 2.66 These two aggregate values together generate the fracture critical state: {aggregate phase difference: 18.7 degrees, aggregate intensity value: 2.66 }.

[0035] Please refer to Figure 5 The specific steps of S4 are as follows: S401: Based on the fracture critical state, match the concentrated frequency section recorded in the energy spectrum data, extract the spectral energy value of the section and perform continuous interval aggregation to obtain the energy concentrated frequency section; Based on the generated fracture critical state {aggregate phase difference: 18.7 degrees, aggregate intensity value: 2.66 }, trace back to the original pressure event that generated the state, and the dominant frequency is concentrated in the range of 18 Hz to 20 Hz. This frequency range is determined as the concentrated frequency section. Then, in the energy spectrum data recorded synchronously with the pressure sensor, all spectral energy values in this section are extracted. Specifically, all energy points with frequency labels between 18 Hz and 20 Hz in the energy spectrum are retrieved, such as 1.12 AU^2 at 18.0 Hz, 2.53 AU^2 at 18.5 Hz, 1.45 AU^2 at 19.0 Hz, 0.75 AU^2 at 19.5 Hz, and 0.52 AU^2 at 20.0 Hz. Subsequently, these extracted spectral energy values are aggregated in continuous intervals, i.e., summed. The total energy is 1.12 + 2.53 + 1.45 + 0.75 + 0.52 = 6.37 AU^2. Finally, this aggregated energy value is combined with the corresponding frequency interval to obtain the energy concentrated frequency section, with the data structure being [18 Hz, 20 Hz], 6.37 AU^2.

[0036] S402: Call the energy concentrated frequency section and combine the time domain sampling points of the crack propagation rate peak signal to perform Fourier transform on the signal to obtain frequency components, and then map the frequency components to the energy concentrated frequency section to generate a frequency conversion result; The generated energy concentration frequency band [18 Hz, 20 Hz] has an energy of 6.37 AU2. Meanwhile, according to the crack propagation rate peak event identified in S203, the sampling point in the original time domain signal is located. Taking the first peak event (time point 1.283 seconds) as an example, a signal segment with a length of 1024 sampling points centered at this time point is cut from the original synthetic strain rate time domain signal. At a sampling frequency of 1000 Hz, the time range covered by this signal segment is from 0.771 seconds to 1.794 seconds. Then, a fast Fourier transform is performed on this signal segment to convert it into frequency components in the frequency domain. After the transformation, a series of frequency points and their corresponding amplitudes are obtained, for example, the amplitude at 18.5 Hz is 0.45 / , and the amplitude at 19.5 Hz is 0.32 / . Then, all frequency components are mapped to the energy concentration frequency band [18 Hz, 20 Hz]. The mapping process is to check whether the frequency value of each frequency component falls within the interval. In this example, both 18.5 Hz and 19.5 Hz are within the interval, so their corresponding frequency components are successfully mapped. Finally, all successfully mapped frequency components are collected to generate the frequency conversion result, which is 18.5 Hz, 0.45 / , 19.5 Hz, 0.32 / .

[0037] S403: Perform resonance judgment according to the frequency conversion result and the numerical interval of the energy concentration frequency band. If the frequency component and the energy concentration frequency band form a resonance relationship, the corresponding time period is recorded as a response section, and a resonance response section is generated. According to the generated frequency conversion result, i.e., 18.5 Hz, 0.45 / , 19.5 Hz, 0.32 / , resonance judgment is performed with the energy concentration frequency band [18 Hz, 20 Hz]. The specific process of the judgment is to check each frequency component in the frequency conversion result to confirm whether its amplitude value exceeds a preset "resonance judgment amplitude threshold value". The threshold value is set through a control experiment of 50 cases of in-vitro perfused pig aortic models. In the experiment, for blood vessels subjected to non-resonant weak mechanical disturbance, the maximum amplitude value of the strain rate signal after Fourier transform at the disturbance frequency is the 99th percentile of 0.25 / , so this value is set as the threshold value. Next, this threshold value is applied to the frequency conversion result of this embodiment. The amplitude of the first frequency component is 0.45 / Greater than the threshold of 0.25 / The resonance condition is met. The amplitude of the second component is 0.32. / Also exceeding this threshold, the resonance condition is satisfied. Since at least one frequency component satisfies the condition, a resonance relationship is determined to have formed within the time period. Subsequently, the start and end times of the signal segment used for Fourier transform, i.e., from 0.771 seconds to 1.794 seconds, are recorded as a response segment. All time periods determined to have formed a resonance relationship are collected to generate the resonance response segment. In this example, the resonance response segment is [0.771 seconds, 1.794 seconds].

[0038] Please see Figure 6 The specific steps of S5 are as follows: S501: Based on the corresponding time index of the resonance response segment and the fracture critical state, the spectral amplitude value recorded in the resonance response segment and the stress intensity recorded in the fracture critical state are weighted and multiplied, and the product results are summed and normalized to obtain the risk index sequence. Based on the generated resonance response segment [0.771 seconds, 1.794 seconds] and the corresponding original event time index of 1.283 seconds, and the fracture critical state generated by S303 {critical phase difference: 18.7 degrees, critical intensity value: 2.66}, The risk index is calculated by first performing a weighted product of the spectral amplitude values ​​recorded within the resonant response region and the stress intensity values ​​recorded in the critical fracture state. The spectral amplitude values ​​are derived from the S402 result, i.e., 18.5 Hz, 0.45 / 19.5 Hz, 0.32 / The stress intensity value uses the critical strength value of 2.66. The calculation involves amplitude and intensity weights, set to 0.4 and 0.6 respectively. These weights are based on a retrospective study of 100 surgeries (50 with bleeding, 50 without), determined through statistical discriminant analysis to assess the contribution of each factor to bleeding risk. For each frequency component within the resonant response range, its amplitude and critical intensity values ​​are multiplied by their respective weights, and then the two products are summed. For the first frequency component, the calculation is 0.45 multiplied by 0.4, plus 2.66 multiplied by 0.6, resulting in 1.776. For the second frequency component, the calculation is 0.32 multiplied by 0.4, plus 2.66 multiplied by 0.6, resulting in 1.724. Subsequently, the calculation results for all frequency components are summed to obtain 3.5. Finally, this sum is normalized. The normalization baseline is the theoretical maximum risk value, based on the maximum resonant amplitude of 1.2 recorded in historical data. / and maximum critical strength 5.0 , set a maximum of 5 resonance components contained in an event calculated value is 17.4. The normalization process is to divide the current calculation of the total risk value 3.5 by the theoretical maximum risk value 17.4, multiplied by 100, the risk index is about 20.11. Repeat this process for each event in the monitoring process, generate a time sequence of risk index sequence, for example, contains (1.283 seconds, 20.11) and another event (2.500 seconds, 18.74).

[0039] S502: call risk index sequence, and interval comparison of risk index value of multi-time point and preset risk level threshold interval, mark all risk index value exceeding risk level threshold interval time point position, generate risk label index set; Call the generated risk index sequence. For full description, set the sequence has been expanded to [(1.283 seconds, 20.11), (2.500 seconds, 18.74), (3.150 seconds, 75.42), (4.020 seconds, 88.16)]. Then, interval comparison of risk index value of each time point in the sequence and preset risk level threshold interval. The setting of risk level threshold interval is based on statistical analysis of 200 cases of historical surgery data, by drawing the receiver operating characteristic curve and using the maximum principle of Youden index to determine, the risk index is divided into three levels: the low risk interval is, the medium risk interval is (30, 70], the high risk interval is (70, 100]. The goal of this step is to mark all time points whose risk index value falls into the high risk interval. The specific operation is to compare each risk index value with the lower limit value 70 of the high risk interval. For time point 1.283 seconds, risk index 20.11 is not greater than 70, not marked. For time point 2.500 seconds, risk index 18.74 is not greater than 70, not marked. For time point 3.150 seconds, risk index 75.42 is greater than 70, the time point is marked. For time point 4.020 seconds, risk index 88.16 is greater than 70, the time point is also marked. Collect all the marked time point positions to generate risk label index set, which is {3.150 seconds, 4.020 seconds} in this embodiment.

[0040] S503: according to the corresponding relationship between risk label index set and risk index sequence, integrate the risk index value of the marked position into the bleeding risk record data frame, and attach the bleeding risk label, generate the bleeding risk result in general surgery resection; Risk level threshold interval is set by stratified statistical analysis of the numerical distribution of risk data frame in the operation process; According to the generated risk label index set {3.150 seconds, 4.020 seconds} and the complete risk index sequence generated by S501, subsequent processing is performed. First, each time point in the risk label index set is traversed, and the risk index value corresponding to the time point is extracted from the risk index sequence. For the time point 3.150 seconds, its risk index value 75.42 is extracted; for the time point 4.020 seconds, its risk index value 88.16 is extracted. Next, the integrated data is attached with the bleeding risk label. Since these points are marked because their risk index values fall into the high-risk interval, the data points are attached with the "high-risk" label. Finally, the integrated data is organized into a bleeding risk record data frame. Each record of the data frame contains information of a high-risk event. For example, one record contains the time point 3.150 seconds, the risk index 75.42, and the risk level "high risk". Another record contains the time point 4.020 seconds, the risk index 88.16, and the risk level "high risk". The data frame containing all high-risk event records is the final generated bleeding risk result in the general surgery resection.

[0041] Referring to Figure 7 , the bleeding risk assessment system in general surgery resection comprises: a pressure analysis module that collects the instantaneous pressure value of the extravascular wall of the operation area in resection and calculates the difference, extracts the main frequency and frequency energy value of the difference result and performs time sequence labeling, generates the pressure main frequency feature and transmits it to the crack feature module; a crack feature module that collects the multi-directional strain rate signal of the extravascular wall based on the time label corresponding to the pressure main frequency feature, and performs vector synthesis as a crack propagation rate curve, extracts the propagation rate peak time point and calculates the intensity level, generates the crack propagation rate peak feature and transmits it to the critical judgment module; a critical judgment module that obtains the corresponding key time points of the crack propagation rate peak feature and the pressure main frequency feature and calculates the phase difference, jointly analyzes the phase difference and the crack propagation rate intensity level, generates the fracture critical state and transmits it to the response recognition module; a response recognition module that matches the energy concentration frequency band based on the fracture critical state, combines the corresponding crack propagation rate curve and performs frequency conversion, and judges whether there is resonance combined with the energy concentration frequency band, generates the resonance response section and transmits it to the risk assessment module; a risk assessment module that performs weighted calculation of the risk index based on the resonance response section and the fracture critical state, and attaches the bleeding risk label if the risk index exceeds the risk level threshold interval, and generates the bleeding risk result in general surgery resection.

[0042] The above merely illustrates the specific embodiments of the present application, but the protection scope of the present application is not limited thereto, any person skilled in the art can easily think of the changes or replacements within the technical range disclosed by the present application, which should be covered in the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A method for assessing the risk of bleeding in a general surgery resection, characterized in that, Includes the following steps: S1: Collect the instantaneous pressure value of the outer wall of the blood vessel in the operation area during the resection and calculate the difference. Extract the main frequency and frequency domain energy value from the difference results and perform time-series calibration to generate pressure main frequency characteristics. S2: Based on the time marker corresponding to the pressure main frequency feature, collect multi-directional strain rate signals of the outer wall of the blood vessel and perform vector synthesis as a crack propagation rate curve, extract the peak time point of the propagation rate and calculate the intensity level to generate the peak feature of the crack propagation rate. S3: Obtain the corresponding key time points of the peak feature of the crack propagation rate and the main frequency feature of the pressure and calculate the phase difference. Perform joint analysis of the phase difference and the crack propagation rate intensity level to generate the critical fracture state. S4: Based on the fracture critical state, match the energy concentration frequency band, combine the corresponding fracture propagation rate curve and perform frequency conversion, and combine the energy concentration frequency band to determine whether resonance occurs, and generate a resonance response segment. S5: Calculate the risk index by weighting the resonant response segment and the critical fracture state. If the risk index exceeds the risk level threshold range, add a bleeding risk label to generate the bleeding risk result during general surgical resection.

2. The method for assessing intraoperative bleeding risk during general surgical resection according to claim 1, characterized in that, The pressure main frequency characteristics include the main frequency value, frequency domain energy distribution, and time calibration information; the crack propagation rate peak characteristics include the propagation rate peak value and intensity level; the fracture critical state includes key time points, phase difference values, and fracture intensity thresholds; the resonance response segment includes frequency conversion results, energy concentration frequency bands, and resonance identification intervals; and the bleeding risk results include risk index values, risk level thresholds, and bleeding risk labels.

3. The method for assessing intraoperative bleeding risk during general surgical resection according to claim 1, characterized in that, The specific steps of S1 are as follows: S101: Collect the instantaneous pressure value of the outer wall of the blood vessel in the operation area during the resection, perform differential calculation on the pressure values ​​of adjacent sampling points, and reassemble them according to the sampling time sequence to obtain the differential pressure sequence; S102: Perform fast Fourier transform operation based on the differential pressure sequence, extract the amplitude peaks in the spectrum distribution and determine the corresponding frequency values, and at the same time perform cumulative calculation on the energy of multiple frequency components to obtain a frequency energy parameter set; S103: Call the frequency energy parameter set, perform corresponding time-series calibration of the main frequency value and frequency domain energy vector according to the time sampling index, and collect the calibrated frequency value as a feature factor to generate pressure main frequency feature.

4. The method for assessing intraoperative bleeding risk during general surgical resection according to claim 1, characterized in that, The specific steps of S2 are as follows: S201: Based on the time stamp corresponding to the main frequency characteristics of the pressure, multi-directional strain rate signals of the outer wall of the blood vessel under the same time index are collected, and the multi-directional signals are decomposed into vectors according to coordinate components. Then, the corresponding components are vector synthesized to obtain the synthesized strain rate curve. S202: Call the synthetic strain rate curve, retrieve the rate values ​​of multiple time sampling points in the curve, extract the peak time point of the expansion rate, calculate the difference intensity with the preset expansion rate benchmark value, and obtain the peak time intensity group. S203: Based on the peak time intensity group, bind the time point index with the intensity difference, and use the bound time point and intensity value as feature output to generate the peak feature of crack propagation rate.

5. The method for assessing intraoperative bleeding risk during general surgical resection according to claim 4, characterized in that, The extended rate benchmark value is set by retrieving all rate values ​​within the sampling interval before and after the time point corresponding to the maximum rate value and then performing an arithmetic average.

6. The method for assessing intraoperative bleeding risk during general surgical resection according to claim 1, characterized in that, The specific steps for S3 are as follows: S301: Obtain the time index of the peak feature of the crack propagation rate and the pressure main frequency feature, retrieve the key sampling points corresponding to the time indexes, normalize and convert the time difference of the corresponding points to obtain the phase angle, and calculate the phase difference sequence in time order. S302: Call the phase difference sequence and pair it point by point with the intensity level recorded in the peak feature of the crack propagation rate. Perform joint distribution calculation on multiple pairs of paired data to obtain the joint distribution of phase intensity. S303: Based on the joint distribution of phase intensity, analyze the phase intensity distribution density of multiple points and combine it with the phase density threshold to screen the critical interval. Aggregate the phase difference and intensity value corresponding to the critical interval to generate the fracture critical state.

7. The method for assessing intraoperative bleeding risk during general surgical resection according to claim 6, characterized in that, The phase density threshold is determined by performing density statistics on the joint distribution results formed after pairing phase difference and intensity level, retrieving the distribution frequency of multiple numerical points and calculating the average frequency, and setting it in combination with the multiple coefficient of the benchmark value.

8. The method for assessing intraoperative bleeding risk during general surgical resection according to claim 1, characterized in that, The specific steps of S4 are as follows: S401: Based on the concentrated frequency segments recorded in the energy spectrum data of the fracture critical state matching, extract the spectral energy values ​​of the segments and perform continuous interval aggregation to obtain the concentrated energy frequency band; S402: Call the energy concentration frequency band and combine it with the time domain sampling points of the peak signal of the fracture propagation rate to perform a Fourier transform on the signal to obtain the frequency component. Then, map the frequency component to the energy concentration frequency band to generate the frequency conversion result. S403: Perform resonance determination based on the frequency conversion result and the numerical range of the energy concentration frequency band. If the frequency component and the energy concentration frequency band form a resonance relationship, record the corresponding time period as a response segment and generate a resonance response segment.

9. The method for assessing intraoperative bleeding risk during general surgical resection according to claim 1, characterized in that, The specific steps of S5 are as follows: S501: Based on the corresponding time index of the resonance response segment and the fracture critical state, perform a weighted product operation on the spectral amplitude value recorded in the resonance response segment and the stress intensity recorded in the fracture critical state, and sum and normalize the product results to obtain the risk index sequence. S502: Call the risk index sequence and compare the risk index values ​​of multiple time points with the preset risk level threshold range, mark the time point positions where the risk index values ​​exceed the risk level threshold range, and generate a risk label index set; S503: Based on the correspondence between the risk label index set and the risk index sequence, the risk index values ​​of the marked positions are integrated into a bleeding risk record data frame, and a bleeding risk label is attached to generate a bleeding risk result during general surgical resection. The risk level threshold range is set by performing hierarchical statistical analysis on the numerical distribution of risk data frames during the surgical procedure.

10. A general surgical resection intraoperative bleeding risk assessment system, characterized in that, The system is used to implement the intraoperative bleeding risk assessment method for general surgical resection as described in any one of claims 1-9, the system comprising: The pressure analysis module collects the instantaneous pressure value of the outer wall of the vessel in the surgical area during resection and calculates the difference. It extracts the main frequency and frequency domain energy value from the difference results and performs time-series calibration, generates pressure main frequency features and transmits them to the rift feature module. The fissure feature module, based on the time marker corresponding to the pressure main frequency feature, collects multi-directional strain rate signals of the outer wall of the blood vessel and performs vector synthesis as a fissure propagation rate curve, extracts the peak time point of the propagation rate and calculates the intensity level, generates the peak feature of the fissure propagation rate and transmits it to the critical determination module. The critical determination module obtains the corresponding key time points of the peak feature of the crack propagation rate and the main frequency feature of the pressure and calculates the phase difference. It then performs joint analysis of the phase difference and the crack propagation rate intensity level to generate the critical fracture state and transmit it to the response identification module. The response identification module matches the energy concentration frequency band based on the fracture critical state, combines the corresponding fracture propagation rate curve and performs frequency conversion, and determines whether resonance occurs based on the energy concentration frequency band, generates a resonance response segment and transmits it to the risk assessment module. The risk assessment module calculates a risk index based on the resonance response segment and the critical fracture state. If the risk index exceeds the risk level threshold range, a bleeding risk label is added to generate a bleeding risk result during general surgical resection.

Images