Method and system for monitoring blood flow state of dorsal foot artery

Abstract

The invention provides a dorsal foot artery blood flow state monitoring method and system, and relates to the technical field of high-end medical instruments and digital diagnosis and treatment, and the method comprises the following steps: collecting a multi-source optical signal of a dorsal foot artery; extracting morphological characteristics of the preprocessed multi-source optical signals; in combination with a pre-trained multiple linear regression model, the morphological features are processed, and local blood pressure is determined; according to the local blood pressure, the stiffness of the peripheral blood vessels and the microvascular dilation function of the dorsal foot artery area are evaluated; according to an evaluation result, determining an automatic adjustment index of the foot dorsum area by constructing a transfer function; fusing the morphological characteristics, the local blood pressure, the automatic adjustment index of the foot dorsal area and the static risk factor of the electronic medical record of the patient through a lightweight gradient boosting tree model, and determining the comprehensive perfusion index of the foot dorsal artery; designing a self-adaptive alarm logic device based on context awareness according to the dorsal foot artery comprehensive perfusion index; and outputting a blood flow state monitoring report according to the self-adaptive alarm logic device.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of high-end medical devices and digital diagnosis and treatment technology, and particularly relates to a dorsalis pedis artery blood flow state monitoring method and system. BACKGROUND

[0002] As an important blood supply vessel of the lower extremity distal, the blood flow state of the dorsalis pedis artery can directly reflect the lower extremity distal perfusion level and peripheral vascular function condition, and has important significance in clinical scenes such as peripheral arterial disease, diabetic foot, and postoperative complication monitoring of vascular intervention. Especially after transfemoral intervention surgery, local blood vessels may cause distal perfusion to decrease due to thrombosis, vasospasm or intimal injury, and the dorsalis pedis artery blood flow change often occurs earlier than obvious clinical symptoms. Therefore, continuous, fine and quantitative monitoring of the dorsalis pedis artery blood flow state is an important technical basis for realizing early identification and intervention of lower extremity ischemia risk.

[0003] Current dorsalis pedis artery blood flow state monitoring methods mostly rely on manual palpation, single Doppler ultrasound detection or static blood flow velocity measurement, and can usually only provide qualitative or semi-quantitative information at a certain time, and it is difficult to reflect the dynamic change characteristics and regulation capacity of blood flow.

[0004] However, the existing dorsalis pedis artery blood flow state monitoring method often lacks joint analysis of local blood pressure, microcirculation blood flow and vascular function state, and it is difficult to realize accurate risk judgment in a complex postoperative situation, and it is highly dependent on the experience of the operator, and is not suitable for long-time continuous monitoring and intelligent early warning demand, which limits its application effect in fine postoperative monitoring and intelligent medical scene. SUMMARY

[0005] In view of the above deficiencies of the prior art, the purpose of the embodiments of the present application is to provide a dorsalis pedis artery blood flow state monitoring method, which can solve the technical problems that the existing dorsalis pedis artery blood flow state monitoring method often lacks joint analysis of local blood pressure, microcirculation blood flow and vascular function state, and it is difficult to realize accurate risk judgment in a complex postoperative situation, and it is highly dependent on the experience of the operator, and is not suitable for long-time continuous monitoring and intelligent early warning demand, which limits its application effect in fine postoperative monitoring and intelligent medical scene.

[0006] The first aspect of the embodiments of the present application provides a dorsalis pedis artery blood flow state monitoring method, comprising:

[0007] S1: acquiring multi-source optical signals of the dorsalis pedis artery through an integrated probe;

[0008] S2: extracting morphological features of the preprocessed multi-source optical signals through a pulse wave analysis method;

[0009] S3: mapping the morphological features to determine the local blood pressure by combining a pre-trained multiple linear regression model;

[0010] S4: evaluating the stiffness of the perivascular and microvascular vasodilation function of the dorsal artery region of the foot according to the local blood pressure and combining the vascular resistance principle;

[0011] S5: constructing a transfer function between the local blood pressure and the blood flow velocity in the multi-source optical signal according to the evaluation result, and determining the dorsal region automatic regulation index through the transfer function;

[0012] S6: fusing the morphological features, the local blood pressure, the dorsal region automatic regulation index and the static risk factors of the patient's electronic medical record through a light gradient boosting tree model to determine the dorsal artery comprehensive perfusion index;

[0013] S7: designing a context-aware adaptive alarm logic for the foot dorsal artery blood flow state monitoring of the patient after the transfemoral intervention according to the dorsal artery comprehensive perfusion index;

[0014] S8: outputting a visual report of the foot dorsal artery blood flow state monitoring result according to the context-aware adaptive alarm logic.

[0015] The second aspect of the embodiment of the application provides a foot dorsal artery blood flow state monitoring system, comprising a processor and a memory.

[0016] The memory stores programs or instructions executable on the processor, and the programs or instructions are executed by the processor to implement the steps of the foot dorsal artery blood flow state monitoring method according to the first aspect.

[0017] The third aspect of the embodiment of the application provides a readable storage medium, and the readable storage medium stores programs or instructions, and the programs or instructions are executed by the processor to implement the steps of the foot dorsal artery blood flow state monitoring method according to the first aspect.

[0018] The technical scheme provided by the embodiment of the application has at least the following beneficial effects:

[0019] In the embodiment of the present application, the multi-source blood flow related signals of the dorsalis pedis artery are integrated, and based on the pulse waveform morphology analysis, local blood pressure inversion and modeling of vascular resistance and automatic regulation capacity, a comprehensive perfusion index of the dorsalis pedis artery is constructed to realize continuous, quantitative and multi-dimensional evaluation of the blood flow state of the dorsalis pedis artery. By fusing the blood flow morphology characteristics, local blood pressure level, automatic regulation index and individual static risk factors of the patient, the real perfusion state and its change trend of the dorsalis pedis artery after operation can be more accurately reflected. Further combined with the adaptive alarm logic of postoperative situation perception, the early identification and graded early warning of the abnormal perfusion of the dorsalis pedis artery are realized, so as to effectively make up for the shortcomings of the existing methods, such as relying on a single index, being difficult to dynamically monitor and early warning, and improve the accuracy, safety and clinical practical value of the lower limb blood flow monitoring after transfemoral intervention. BRIEF DESCRIPTION OF DRAWINGS

[0020] The accompanying drawings are included to provide a further understanding of the present application, and are incorporated in and constitute a part of this specification. The drawings illustrate embodiments of the present application and, together with the description, serve to explain the principles of the present application. In the drawings:

[0021] Figure 1 FIG. 1 is a flowchart of a dorsalis pedis artery blood flow state monitoring method according to an embodiment of the present application.

[0022] Figure 2 FIG. 2 is a structural diagram of a dorsalis pedis artery blood flow state monitoring system according to an embodiment of the present application. DETAILED DESCRIPTION

[0023] In order for those skilled in the art to better understand the technical solutions in the embodiments of the present application, the technical solutions of the present application will be described clearly and completely below in conjunction with the drawings. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments of the present application. It should be understood that these descriptions are only exemplary and are not used to limit the scope of the present application. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without creative labor should be within the scope of protection of the present application.

[0024] The dorsalis pedis artery blood flow state monitoring method provided by the embodiments of the present application will be described in detail below in conjunction with the drawings and specific embodiments and application scenarios.

[0025] Reference is made to the accompanying drawings Figure 1 FIG. 1 is a flowchart of a dorsalis pedis artery blood flow state monitoring method according to an embodiment of the present application.

[0026] The embodiment of the present application provides a dorsalis pedis artery blood flow state monitoring method, which can comprise the following steps:

[0027] S1: Collecting multi-source optical signals of the dorsalis pedis artery through an integrated probe.

[0028] Among them, the integrated probe refers to a collection device that integrates multiple non-invasive blood flow detection units in an integrated structure in the same probe, which is used to synchronously acquire the dorsalis pedis artery related signals at the same measurement position and the same time scale.

[0029] Specifically, the integrated probe comprises a miniaturized Doppler ultrasound unit, a dual-channel photoplethysmography sensor and a diffuse correlation spectroscopy (DCS) or laser Doppler flowmetry (LDF) module.

[0030] Among them, the multi-source optical signals refer to optical or photoelectric signals related to the blood flow state of the dorsalis pedis artery, which are obtained through different physical mechanisms.

[0031] In a possible implementation, the multi-source optical signals comprise blood flow velocity of the main stem of the dorsalis pedis artery, microvascular pulse waveform and microcirculation blood flow in the dorsalis pedis region.

[0032] It should be noted that by using the integrated probe to synchronously collect the multi-source optical signals of the dorsalis pedis artery, the macroscopic blood flow characteristics and microcirculation changes can be cooperatively acquired without increasing the trauma and operation complexity of the patient, thereby effectively avoiding the time asynchronization and measurement deviation problems caused by multi-device time-sharing measurement, and improving the consistency and reliability of the signals. At the same time, the multi-source optical signals have complementarity in physical mechanism and information dimension, and through fusion analysis, the sensitivity and robustness to the changes in the blood flow state of the dorsalis pedis artery can be significantly enhanced, thereby providing a stable and rich data basis for subsequent local blood pressure inversion, blood vessel function evaluation and intelligent early warning, and improving the engineering implementability and clinical application value of the monitoring system as a whole.

[0033] S2: Extracting morphological features of the preprocessed multi-source optical signals through a pulse wave analysis method.

[0034] Among them, the pulse wave analysis method refers to an analysis method that takes a cardiac cycle as a basic analysis unit, segments, aligns and identifies feature points of the collected pulse wave or blood flow related signals, and is used to characterize the periodic variation characteristics of blood flow caused by heartbeats.

[0035] Among them, the morphological features refer to feature parameters extracted from the pulse waveform, which can reflect the characteristics of blood vessel compliance, resistance and hemodynamic state, including pulse wave rising time, main wave peak value characteristics, double pulse wave saliency and time relationship of key feature points.

[0036] Specifically, the preprocessing includes: performing spectrum analysis on the microcirculation blood flow of the dorsal foot area to extract a pulsatility index and a resistance index.

[0037] The microvascular pulse waveform is filtered and processed to eliminate motion artifacts to extract a clear single pulse period waveform.

[0038] The microcirculation blood flow of the dorsal foot area is denoised and smoothed.

[0039] It should be noted that by using the pulse wave analysis method to extract morphological features from the preprocessed multi-source optical signals, the originally continuous and complex blood flow signals can be converted into feature parameters with clear physiological significance and quantifiable comparison value, thereby reducing signal redundancy and highlighting information closely related to blood flow state. This method not only reflects the dynamic change characteristics of the dorsal foot artery in different cardiac cycles, but also maintains high sensitivity to changes in vascular elasticity and abnormal microcirculation function, providing clear and stable feature input for subsequent local blood pressure estimation, vascular function evaluation, and comprehensive perfusion discrimination, effectively improving the accuracy and interpretability of the overall monitoring method.

[0040] In one possible implementation, S2 specifically includes:

[0041] S201: According to the trough position of the microvascular pulse waveform in the preprocessed multi-source optical signal, the microvascular pulse waveform is divided into multiple single cardiac cycle pulses.

[0042] Wherein, the single cardiac cycle pulse refers to a pulse wave segment corresponding to one complete heart beat formed by taking adjacent troughs as segmentation points.

[0043] S202: In the monitoring time window, an epoch pulse waveform is constructed for each single cardiac cycle pulse.

[0044] Wherein, the epoch pulse waveform refers to a representative pulse waveform obtained by amplitude normalization, time alignment and point-by-point averaging of multiple single cardiac cycle pulses with complete morphology and no obvious artifacts in the same monitoring time window.

[0045] Specifically, by performing amplitude normalization and time alignment on multiple single cardiac cycle pulses with complete morphology and no artifacts in each monitoring time window, and performing point-by-point averaging, an epoch pulse waveform representing the time window is obtained.

[0046] S203: By pulse wave analysis method, key feature points of the epoch pulse waveform and time stamps corresponding to the key feature points are extracted.

[0047] Wherein, the key features include: the starting point of the pulse wave, the vertex of the systolic main wave peak, the inflection point of the double wave, and the vertex of the diastolic double wave peak.

[0048] S204: Constructing morphological features according to the key feature points and the time stamps corresponding to the key feature points.

[0049] It should be noted that by first accurately dividing the microvessel pulse waveform into multiple single cardiac cycle pulses based on the trough position, and constructing the epoch pulse waveform within the monitoring time window, the interference of random noise, motion artifacts and occasional abnormal pulses on the analysis results can be effectively suppressed, making the obtained pulse waveform more stable and representative. On this basis, by extracting key feature points with clear physiological significance and their time stamps to construct morphological features, not only the core kinetic information of the dorsalis pedis artery blood flow waveform is retained, but also the robustness and repeatability of the features are significantly improved, thereby providing a more reliable and structured data basis for subsequent local blood pressure estimation and vascular function evaluation, which is conducive to long-term continuous monitoring and intelligent analysis.

[0050] S3: Mapping processing of morphological features is performed in combination with a pre-trained multiple linear regression model to determine the local blood pressure.

[0051] The multiple linear regression model refers to a statistical regression model established by taking multiple morphological features as independent variables and local blood pressure as dependent variable, which is used to depict the linear mapping relationship between pulse waveform morphological features and blood pressure parameters.

[0052] The local blood pressure refers to the blood pressure parameter corresponding to the dorsalis pedis artery region, including systolic pressure and diastolic pressure, which is used to represent the real blood flow hemodynamic pressure state of the region.

[0053] It should be noted that by introducing the pre-trained multiple linear regression model to process the morphological features of the dorsalis pedis artery pulse waveform, continuous and non-invasive estimation of the local blood pressure of the dorsalis pedis artery can be achieved without the need for additional inflatable cuffs or invasive sensors, thereby significantly improving the comfort and operability of the monitoring process. At the same time, the multiple linear regression model has a simple structure and high computational efficiency, which is suitable for real-time operation in resource-limited monitoring terminals or wearable devices, which is conducive to long-term stable monitoring. In addition, modeling based on the mapping relationship between morphological features and blood pressure makes the obtained local blood pressure results have clear physiological interpretation basis, which provides key and reliable core parameter support for subsequent vascular resistance evaluation, automatic regulation ability analysis and comprehensive perfusion discrimination.

[0054] In one possible implementation, the training process of the pre-trained multiple linear regression model in S3 specifically includes:

[0055] S301: Combine morphological features and reference local blood pressure to construct a training sample set, which includes: multiple sets of records and dorsolateral foot epoch samples. The multiple sets of records include multiple dorsolateral foot epoch samples, and the dorsolateral foot epoch samples include morphological feature vectors and corresponding reference local blood pressure labels.

[0056] The training sample set refers to the dataset used to establish the mapping relationship between morphological features and local blood pressure. It consists of multiple independent records, each containing multiple dorsolateral foot epoch samples. A dorsolateral foot epoch sample is a representative pulse wave analysis unit formed within a fixed monitoring time window, which includes the corresponding morphological feature vector and a synchronously obtained reference local blood pressure label.

[0057] Among them, reference local blood pressure refers to the dorsalis pedis artery blood pressure reference value obtained through standard blood pressure measurement methods and used for model training and calibration.

[0058] S302: Input the training sample set into the multiple linear regression model, and output the initial parameter vector:

[0059]

[0060] in, This represents the initial parameter vector of the multiple linear regression model obtained in the first iteration, where DBP represents diastolic blood pressure and SBP represents systolic blood pressure. Minimization means finding the parameter vector that minimizes the sum of squared errors. , This represents the PPG morphological feature vector of the i-th epoch. (1) This indicates that it comes from the first set of records. This indicates the predicted local blood pressure value. Let represent the reference local blood pressure in the i-th epoch, where i = 1, ..., n, and n represents the total number of epochs.

[0061] S303: Calculate the individualized offsets of systolic and diastolic blood pressure based on the initial parameter vector and reference local blood pressure.

[0062]

[0063] in, This represents the individualized offset between systolic and diastolic blood pressure. This represents the morphological feature vector corresponding to the i-th epoch. This represents the reference local blood pressure at the i-th epoch, where i = 1, ..., m, and m represents the number of epochs used to calculate the offset.

[0064] The individualized offset refers to a compensation term calculated for systematic errors introduced by different individuals due to differences in blood vessel conditions, sensor wearing positions, and other factors.

[0065] S304: The morphological features and their corresponding reference local blood pressure data used for model parameter updating are compensated for the individualized offset until the estimated bias of the multiple linear regression model on each group of records remains stable, and the final parameter vector of the multiple linear regression model is determined:

[0066]

[0067] wherein, represents the final parameter vector of the multiple linear regression model.

[0068] S305: According to the final parameter vector, the pre-trained multiple linear regression model is determined.

[0069] It should be noted that by constructing a training sample set containing multiple groups of dorsum pedis time series samples and introducing an individualized offset compensation mechanism in the model training process, the physiological differences and measurement condition differences between different individuals can be effectively absorbed while keeping the multiple linear regression model structure simple, significantly improving the accuracy and consistency of local blood pressure estimation. Through iterative updating of model parameters combined with multiple records, the model can maintain stable estimation performance under different data distribution conditions, avoiding fitting bias caused by a single data source or short-term sampling. This training method takes into account both global statistical rules and individual difference correction, providing a robust, interpretable, and easy-to-implement model foundation for subsequent real-time blood pressure estimation, which is beneficial to the promotion and application in actual clinical monitoring environment.

[0070] S4: According to the local blood pressure, the stiffness of the surrounding blood vessels and the microvascular dilation function of the dorsum pedis arterial region are evaluated based on the principle of vascular resistance.

[0071] The principle of vascular resistance refers to the basic physical principle for describing the degree of obstruction of blood vessels to blood flow based on the relationship between pressure difference, blood flow, and vascular resistance in hemodynamics.

[0072] The surrounding blood vessels refer to the small and medium-sized arterial and capillary blood vessel network located around the dorsum pedis artery and its branches, which directly participates in the distribution of peripheral blood.

[0073] The stiffness refers to the ability of the blood vessel wall to deform under the action of blood pressure change, which reflects the level of blood vessel elasticity and compliance.

[0074] The microvascular dilation function refers to the ability of microvessels to expand and adjust to changes in blood flow demand under physiological regulation, which is an important indicator for evaluating the health status of peripheral perfusion and microcirculation.

[0075] It should be noted that by evaluating the stiffness of the perivascular vessels and the microvascular diastolic function in the dorsal foot artery region based on local blood pressure and combined with the principle of vascular resistance, the vascular function status of the dorsal foot region can be quantitatively analyzed from the hemodynamic mechanism level, rather than just staying on the surface representation of blood flow velocity or waveform changes. This approach helps to distinguish between passive blood flow fluctuations caused by blood pressure changes and active regulation abnormalities caused by impaired vascular function, thereby more accurately identifying early pathological changes such as microcirculation disorders and decreased vascular elasticity. By introducing vascular resistance-related indicators, key functional parameter support is provided for subsequent automatic regulation capability calculation and comprehensive perfusion state discrimination, significantly improving the diagnostic depth and clinical application value of dorsal foot artery blood flow state monitoring.

[0076] In one possible implementation, S4 specifically includes:

[0077] S401: According to the local blood pressure, the local mean arterial pressure of the dorsal foot artery region is calculated:

[0078]

[0079] Wherein, MAP represents the local mean arterial pressure.

[0080] Wherein, the local mean arterial pressure (MAP) refers to the average effective perfusion pressure acting on the dorsal foot artery and its downstream vascular system within a cardiac cycle, which is calculated by the local systolic pressure (SBP) and the local diastolic pressure (DBP) in a weighted manner.

[0081] S402: According to the principle of vascular resistance, the local mean arterial pressure and the microcirculation blood flow of the dorsal foot region are combined to calculate the perivascular resistance:

[0082]

[0083] Wherein, CVR represents the perivascular resistance, and CBF represents the microcirculation blood flow of the dorsal foot region.

[0084] Wherein, the perivascular resistance (CVR) refers to the comprehensive resistance degree of the small vessels and microvessels in the dorsal foot artery region to blood flow.

[0085] S403: According to the change of the perivascular resistance, the stiffness of the perivascular vessels and the microvascular diastolic function are evaluated:

[0086]

[0087] Wherein, represents the incremental change.

[0088] S5: constructing a transfer function between the local blood pressure and the blood flow velocity in the multi-source optical signal according to the evaluation result, and determining an automatic regulation index of the dorsal foot region through the transfer function.

[0089] wherein the blood flow velocity refers to a dynamic parameter reflecting the speed of blood flow in the dorsal foot artery or a measurement section thereof extracted from the multi-source optical signal. The transfer function refers to a mathematical model used to describe how changes in an input signal affect the response of an output signal in system analysis theory, which is used to characterize the dynamic relationship between changes in the local blood pressure and the response of the blood flow velocity in this step. The automatic regulation index of the dorsal foot region refers to a quantitative index comprehensively reflecting the inhibition ability and response efficiency of the vascular system of the dorsal foot region to pressure disturbance, and is used to represent the physiological regulation ability of the local blood vessels to maintain relatively stable blood flow under different pressure conditions.

[0090] In one possible implementation, S5 specifically includes:

[0091] S501: constructing an input signal sequence and an output signal sequence according to the local mean arterial pressure and the blood flow velocity.

[0092] wherein the input signal sequence is a time sequence of the local mean arterial pressure, and the output signal sequence is a time sequence of the blood flow velocity.

[0093] Specifically, when the measurement section is fixed and the diameter of the tube is assumed to change slowly, the blood flow velocity can be used as a proportional quantity of the blood flow for frequency domain analysis and relative change analysis.

[0094] S502: performing Fourier transform on the input signal sequence and the output signal sequence:

[0095]

[0096] wherein H d (s) represents an impulse response function of the local blood pressure and the blood flow system in the complex frequency domain, s represents a complex frequency domain variable, L( ) represents a Laplace transform operator, and h d (t) represents an impulse response function of the local blood pressure and the blood flow system in the time domain.

[0097] S503: calculating a transfer function between the local blood pressure and the blood flow velocity in the multi-source optical signal based on the Fourier transform result:

[0098]

[0099] wherein, represents a frequency domain transfer function between the local blood pressure and the blood flow velocity, represents a gain (a ratio of an output amplitude to an input amplitude) of the transfer function at a frequency of the transfer function, and denotes the phase (phase lag or lead of the output relative to the input) of the transfer function at frequency denotes the Fourier transform result of the input signal (local mean arterial pressure).

[0100] S504: Calculate the gain and phase characteristics of the transfer function in a specific frequency band:

[0101]

[0102]

[0103] wherein, denotes the average gain characteristic of the transfer function in the low frequency band, denotes the average phase characteristic of the transfer function in the low frequency band, denotes a specific low frequency analysis frequency band, denotes the amplitude of the transfer function at frequency denotes the phase angle of the transfer function at frequency denotes the phase angle of the transfer function at frequency denotes the differential symbol.

[0104] wherein, reflects the overall response amplitude of the blood flow velocity to the local blood pressure change in the range of the low frequency pressure fluctuation concerned, the greater the value, the stronger the "transmission" of the blood flow to the pressure change; the smaller the value, the stronger the buffering ability of the blood vessel to the low frequency pressure fluctuation.

[0105] Specifically, , .

[0106] Specifically, by integrating and averaging the amplitude function and the phase function of the local blood pressure-blood flow velocity transfer function in the preset low frequency band, the low frequency band average gain and average phase characteristics are obtained, which are used to represent the dynamic response ability and automatic regulation characteristics of the dorsal foot region vascular system under slow varying pressure disturbance.

[0107] S505: According to the gain and phase characteristics, combined with the rate index, calculate the dorsal foot region automatic regulation index:

[0108]

[0109]

[0110] wherein, denotes the dorsal foot region automatic regulation index, denotes the weight coefficient of the inverse term of the low frequency transfer gain, denotes the dorsal foot region automatic regulation rate index, denotes the weight coefficient of the dorsal foot region automatic regulation rate index, represents the proportion of change of the vascular resistance in the dorsal foot area relative to its baseline value, represents the proportion of change of the local mean arterial pressure in the dorsal foot area relative to the baseline state, represents the length of the adjustment time window.

[0111] Specifically, by weighting and fusing the inverse of the low-frequency transfer gain (reflecting the "adjustment strength") and the automatic adjustment rate index RoR (reflecting the "adjustment speed"), the dorsal foot area automatic adjustment index is constructed, so as to simultaneously reflect the inhibition ability and response efficiency of the dorsal foot blood vessel system to pressure changes in a unified index.

[0112] It should be noted that by constructing the transfer function between the local blood pressure and the blood flow velocity, and on this basis determining the dorsal foot area automatic adjustment index, the adjustment ability of the dorsal foot artery blood vessel system can be quantitatively analyzed from the perspective of dynamic system, and is no longer limited to the comparison of static blood flow or instantaneous pressure level. This method can simultaneously reflect the buffering ability of the blood vessels to slow-changing pressure disturbance and the response speed to blood flow change, so as to more comprehensively reveal the complete characteristics of the automatic adjustment function of the dorsal foot area blood vessels. By converting the blood pressure-blood flow coupling relationship into a unified index, not only the stability and comparability of the evaluation results are improved, but also key input parameters with clear physiological significance and engineering realizability are provided for subsequent comprehensive perfusion discrimination and situation awareness alarm.

[0113] S6: By a lightweight gradient boosting tree model, the morphological features, the local blood pressure, the dorsal foot area automatic adjustment index, and the static risk factors of the patient's electronic medical record are fused to determine the dorsal artery comprehensive perfusion index.

[0114] Among them, the lightweight gradient boosting tree model refers to an ensemble learning model constructed by multiple weak decision trees through step-by-step iteration, which has relatively low demand for computing resources and storage resources while maintaining high prediction accuracy, and is suitable for deployment in real-time monitoring systems.

[0115] Among them, the static risk factors of the patient's electronic medical record refer to the basic information and past risk information of the patient that remain relatively stable within the current monitoring period, including age, history of underlying diseases, past vascular lesions, etc.

[0116] Among them, the dorsal artery comprehensive perfusion index refers to a quantitative index for reflecting the overall perfusion state of the dorsal artery under the joint action of multi-dimensional blood flow, physiological and individual risk information.

[0117] In one possible implementation, S6 specifically includes:

[0118] S601: Construct an epoch-level fusion feature input vector based on morphological features, local blood pressure, foot dorsum area automatic adjustment index, and static risk factors in the patient's electronic medical record.

[0119] S602: Merge feature subsets with mutual exclusion in the epoch-level fusion feature input vector through an EFB mechanism to generate bound features.

[0120] Wherein, the EFB mechanism refers to a mutual exclusion feature binding mechanism, which is used to merge feature subsets that cannot be activated or valued at the same time in the same epoch, to reduce feature dimension and reduce redundancy.

[0121] Wherein, the bound feature refers to a new feature formed by integrating multiple mutually exclusive original features through the EFB mechanism.

[0122] S603: Calculate the bin value of the bound feature:

[0123]

[0124] Wherein, represents the final discrete bin value of the i-th epoch sample on the bound feature, F(j) represents the j-th original feature item participating in mutual exclusion feature binding, represents the discrete bin value of the i-th epoch sample, represents the starting position of the bin offset interval of the j-th feature item in the bound feature.

[0125] Wherein, the bin value refers to the discrete value obtained after interval division of continuous or discrete features.

[0126] S604: Construct a histogram statistic according to the bound feature and the bin value corresponding to the bound feature.

[0127] Specifically, for the current epoch sample set usedRows, the discrete bin representation of the bound feature is traversed one by one, and the sample falling into the bin is executed in each bin. The statistic accumulation operation is performed to construct the corresponding feature histogram. Specifically, for the j-th sample in the sample set, first determine its belonging bin index, and add the value of the target variable and the sample count in the bin corresponding to the histogram unit. By executing the above accumulation operation on all samples, the cumulative value of the target variable and the sample number in each bin are obtained respectively, thereby forming a histogram structure reflecting the statistical distribution characteristics of different feature value intervals. The histogram is the basic statistical input for subsequent decision tree node discrimination and splitting calculation, which supports the fast and low complexity model inference process.

[0128] Wherein, the histogram is the basic statistical input for subsequent decision tree node discrimination and splitting calculation.

[0129] S605: Calculate the multiple variance gain values corresponding to the candidate split points according to the histogram statistics:

[0130]

[0131] wherein, represents the variance gain value corresponding to the candidate split point d when the binding feature j is used to distinguish, on the current node sample set O, represents the total number of samples in the sample set O, x ij represents the value (bin value) of the i-th epoch sample on the binding feature j, d represents the candidate split point currently considered, represents the gradient value corresponding to the i-th epoch sample under the current model state, x i represents the fusion feature sample of the i-th epoch, represents the number of samples in the sample set O that satisfy , i.e., the number of left child node samples, represents the number of samples in the sample set O that satisfy , i.e., the number of right child node samples.

[0132] wherein, the variance gain value refers to an evaluation index for measuring the distinguishing ability of a certain feature and its candidate split point on the target variable at the current node.

[0133] S606: Introduce a sample weight correction mechanism based on gradient size to correct the variance gain value of the candidate split point:

[0134]

[0135] wherein, represents the result of gradient weight correction on the variance gain value of the candidate split point d, n represents the total number of samples participating in the current calculation, represents the sample subset in the sample set with larger gradient amplitude that is allocated to the left child node, represents the sample subset in the randomly selected sample set among the remaining samples that is allocated to the left child node, a represents the proportion of set A in all samples, and b represents the proportion of set B in all samples, represents the sample subset in the sample set with larger gradient amplitude that is allocated to the right child node, represents the sample subset in the randomly selected sample set among the remaining samples that is allocated to the right child node, represents the number of samples in the left child node under the candidate split point d, represents the number of samples in the right child node under the candidate split point d.

[0136] Specifically, according to the gradient amplitude of the current epoch sample in the deployed light gradient boosting tree model, the sample is divided into a first sample set with large gradient amplitude and a second sample set selected randomly, and a weight compensation factor is applied to the gradient value in the second sample set, so as to obtain a corrected variance gain value.

[0137] S607: Determine the optimal discriminant output of the bound feature according to the corrected variance gain value.

[0138] Specifically, the maximum value of the plurality of variance gain values is taken as the optimal discriminant output.

[0139] S608: Based on the optimal discriminant output, perform forward discriminant calculation on the current epoch-level fusion feature input vector through the light gradient boosting tree model to determine the dorsalis pedis artery comprehensive perfusion index.

[0140] The dorsalis pedis artery comprehensive perfusion index is used to represent the overall perfusion state of the current dorsalis pedis artery under the combined action of morphological features, local blood pressure level, automatic regulation ability and individual static risk factors. The dorsalis pedis artery comprehensive perfusion index is used as the core criterion input of the subsequent context-aware adaptive alarm logic.

[0141] It should be noted that by introducing epoch-level fusion feature modeling, mutually exclusive feature binding and efficient splitting calculation mechanism based on histogram, the feature dimension and computational complexity can be significantly reduced while ensuring the model expression ability, so that complex multi-source physiological information can be efficiently fused in the light gradient boosting tree framework. The gradient weight correction-based variance gain calculation method further enhances the recognition ability of key epoch samples and important physiological changes, improves the sensitivity and stability of the model in the early stage of perfusion abnormalities, and maps multi-dimensional hemodynamic information and individual static risk factors into a single dorsalis pedis artery comprehensive perfusion index, making the monitoring result more intuitive, interpretable and easy to apply clinically, and providing a reliable and robust core criterion input for the subsequent context-aware alarm logic.

[0142] S7: According to the dorsalis pedis artery comprehensive perfusion index, design a context-aware adaptive alarm logic for the dorsalis pedis artery blood flow state monitoring of patients after transfemoral intervention.

[0143] Wherein, context awareness refers to the system's ability to combine postoperative time stages, physiological state changes and individual baseline characteristics to comprehensively identify the current monitoring environment and risk background.

[0144] Wherein, the adaptive alarm logic refers to a logic discriminant module that can dynamically adjust the risk determination conditions and alarm thresholds according to different monitoring contexts, and output graded alarm results.

[0145] In one possible implementation, S7 specifically includes:

[0146] S701: Combine the postoperative time factor and the dorsalis pedis artery comprehensive perfusion index to construct a postoperative monitoring situation state vector.

[0147] The postoperative time factor refers to the time information of the patient after the completion of the femoral artery intervention, and is used to reflect the physiological risk characteristics corresponding to different postoperative stages. The postoperative monitoring situation state vector refers to the state description formed by combining the postoperative time factor and the dorsalis pedis artery comprehensive perfusion index, and is used to represent the clinical situation in which the current monitoring is located.

[0148] S702: Set different reference perfusion thresholds for different postoperative stages according to the postoperative monitoring situation state vector.

[0149] Specifically, the monitoring process after the femoral artery intervention is divided into different stages according to the postoperative time, and the corresponding dorsalis pedis artery comprehensive perfusion index reference threshold is set for each stage. When the postoperative time is in the early high-risk stage of 0-6 hours, a higher reference threshold is used to improve the sensitivity of the system to slight perfusion decline; when the postoperative time is in the recovery transition stage of 6-24 hours, a medium-level reference threshold is used to balance false alarm control and risk identification; when the postoperative time exceeds 24 hours and enters the relatively stable stage, a lower reference threshold is used to reflect the physiological adaptation characteristics after the blood flow is gradually stabilized. Through the above-mentioned dynamic switching of the reference threshold according to the postoperative stage, the same dorsalis pedis artery comprehensive perfusion index has differentiated risk judgment standards in different postoperative stages, thereby realizing the situation perception and adaptive evaluation of the dorsalis pedis blood flow state of the patient after the femoral artery intervention.

[0150] It should be noted that the size of the reference perfusion threshold can be set by the person skilled in the art according to actual needs, which is not limited in the present application.

[0151] S703: According to the reference perfusion threshold, calculate the change rate of the dorsalis pedis artery comprehensive perfusion index within the sliding time window.

[0152] S704: Determine the normalized trend index of the dorsalis pedis artery comprehensive perfusion index according to the change rate.

[0153] Specifically, the change rate is normalized with the baseline perfusion level of the patient individual to obtain the normalized trend index This eliminates the influence of differences in initial perfusion levels among different patients on trend assessment. In this way, the system no longer relies solely on the absolute value of the perfusion index for risk determination, but focuses on characterizing the trend of perfusion status relative to the individual baseline. This allows for the early identification of persistently declining or rapidly deteriorating perfusion risks before the dorsalis pedis artery perfusion falls significantly below a fixed threshold. This normalized trend modeling method combines "rate of change" with "individual baseline," enabling early perception and individualized assessment of postoperative blood flow abnormalities, demonstrating significant innovation and clinical practical value.

[0154] S705: Based on the normalized trend index and combined with the foot dorsum region auto-adjustment index, a multi-condition risk discrimination function is constructed:

[0155]

[0156] in, I represents the value of the multi-condition risk discrimination function at time t, and I() represents the indicator function. This represents the dorsalis pedis artery perfusion index at time t. Indicates the baseline perfusion threshold. This represents the normalized trend index at time t. Indicates the trend discrimination threshold. This represents the automatic adjustment index of the dorsum of the foot at time t. Indicates the threshold for automatic adjustment capability. This represents the weighting coefficient for abnormal perfusion levels. The weighting coefficient indicates an abnormal downward trend in perfusion. This represents the weighting coefficient indicating impaired automatic adjustment capability.

[0157] Among them, the multi-condition risk discrimination function refers to a function model that judges risk by comprehensively considering multiple conditions such as abnormal perfusion level, abnormal perfusion change trend, and impaired automatic adjustment capability.

[0158] S706: Based on the multi-condition risk discrimination function, design a context-aware adaptive alarm logic unit.

[0159] Specifically, based on the calculation results of the risk discrimination function, the system generates a hierarchical adaptive alarm level according to the segmented mapping rule. When =0, it indicates that no risk triggering conditions are met at the current monitoring time, the dorsalis pedis artery perfusion status is within a stable range, corresponding to Level 0 normal state, and no alarm is triggered; when When =1, it indicates the presence of a single risk signal (such as the perfusion index falling below the stage threshold or showing an unfavorable trend), and the system enters a Level 1 warning state to indicate the need for enhanced observation and continuous monitoring; when = 2, indicating that two types of risk factors are detected simultaneously, for example, perfusion decreases and is accompanied by weakened automatic regulation ability, the system enters Level 2 warning state, and prompts that there may be early signs of postoperative complications or limited blood flow in the dorsal foot; when 3, indicating that multiple high-risk conditions exist simultaneously, the perfusion state of the dorsal foot artery has been significantly abnormal, the system enters Level 3 critical state, and triggers the highest level of alarm to prompt immediate clinical intervention; by corresponding the value of the risk discrimination function to the alarm level, the transition from single index triggering to multi-condition fusion and progressive adaptive alarm is realized, so that the alarm result is clear in value and clinically interpretable.

[0160] It should be noted that by incorporating the postoperative time factor, the comprehensive perfusion index of the dorsal foot artery, the normalized trend index, and the automatic regulation index into the multi-condition risk discrimination function, the present application can realize situational awareness and phased adaptive evaluation of the blood flow state of the dorsal foot artery, so that the same perfusion index has differentiated risk determination standards in different postoperative stages, which are consistent with physiological reality; this method not only focuses on whether the perfusion level is below the threshold, but also can identify potential risks of continuous decline or rapid deterioration in advance through trend modeling, thereby significantly improving the early warning ability of postoperative blood flow abnormalities; through multi-condition fusion and graded alarm design, the alarm result not only has clear numerical basis, but also has good clinical interpretability, effectively reducing false positives and false negatives, and improving the safety and practical value of postoperative monitoring of the blood flow of the dorsal foot artery after femoral artery intervention.

[0161] S8: According to the adaptive alarm logic based on situational awareness, output a visual report of the monitoring results of the blood flow state of the dorsal foot artery.

[0162] The visual report of the monitoring results of the blood flow state of the dorsal foot artery refers to an output result that visually displays the blood flow state of the dorsal foot artery and its change process in the form of graphs, curves, index values, and alarm identifiers, etc., for providing clear and easy-to-understand monitoring information to medical staff.

[0163] In one possible implementation, the report of the monitoring results of the blood flow state of the dorsal foot artery specifically includes: the comprehensive perfusion index of the dorsal foot artery and its change trend over time, the local blood pressure parameters and hemodynamic parameters corresponding to the comprehensive perfusion index, the automatic regulation ability evaluation results of the dorsal foot region, the alarm level generated by the adaptive alarm logic based on situational awareness and its triggering basis, and the baseline threshold information and individualized reference information corresponding to the postoperative stage after femoral artery intervention.

[0164] It should be noted that the visualized dorsalis pedis artery blood flow state monitoring result report generated by the context-aware adaptive alarm logic can present the complex multi-dimensional hemodynamic analysis result in an intuitive and structured form, so that medical staff can quickly grasp the dorsalis pedis artery perfusion state and risk change trend without interpreting complex models. This approach helps to improve the information transmission efficiency and decision response speed in postoperative monitoring, while reducing the clinical risks caused by data dispersion or understanding bias. By integrating the monitoring results, alarm levels and their determination basis into the same visualized report, the system's explainability and traceability are further enhanced, and the usability and promotion value of the dorsalis pedis artery blood flow state monitoring in actual clinical application are improved.

[0165] In the embodiment of the present application, by integrating the multi-source blood flow related signals of the dorsalis pedis artery, and based on pulse waveform morphology analysis, local blood pressure inversion and vascular resistance and automatic regulation capacity modeling, a comprehensive perfusion index of the dorsalis pedis artery is constructed to realize continuous, quantitative and multi-dimensional evaluation of the blood flow state of the dorsalis pedis artery. By fusing the blood flow morphology characteristics, local blood pressure level, automatic regulation index and patient individual static risk factors, the real perfusion state of the dorsalis pedis artery and its change trend can be more accurately reflected. Further combined with the postoperative context-aware adaptive alarm logic, early identification and graded early warning of the dorsalis pedis artery perfusion abnormalities are realized, thereby effectively making up for the shortcomings of the existing methods in relying on a single index, being difficult to dynamically monitor and early warning, and improving the accuracy, safety and clinical practical value of post-femoral artery intervention lower limb blood flow monitoring.

[0166] Referring to the accompanying drawings Figure 2 , a structural schematic diagram of a dorsalis pedis artery blood flow state monitoring system provided by an embodiment of the present application is shown.

[0167] An embodiment of the present application provides a dorsalis pedis artery blood flow state monitoring system 20, comprising a processor 201 and a memory 202.

[0168] The memory 202 stores programs or instructions executable on the processor 201, and the programs or instructions are executed by the processor 201 to realize the steps of the above-mentioned dorsalis pedis artery blood flow state monitoring method, and achieve the same technical effects. To avoid repetition, the present application will not be described again.

[0169] It is to be understood that the processor 201 in the embodiments of the present application can be a central processing unit (CPU), and can also be other general-purpose processors, digital signal processors (DSP), application specific integrated circuits (ASIC), field programmable gate arrays (FPGA) or other programmable logic devices, discrete gates or transistor logic components, discrete hardware components, etc. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor.

[0170] It is also to be understood that the memory 202 in the embodiments of the present application can be a volatile memory or a non-volatile memory, or can include both volatile and non-volatile memory. Among them, the non-volatile memory can be a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM) or a flash memory. The volatile memory can be a random access memory (RAM) used as an external cache. By way of example, but not limitation, many forms of random access memory are available, such as static random access memory (SRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (SDRAM), double data rate synchronous dynamic random access memory (DDR SDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchronous link dynamic random access memory (SLDRAM) and direct memory bus random access memory (DRAM).

[0171] The above-described embodiments can be implemented in whole or in part by software, hardware (e.g., circuitry), firmware, or any combination of the three. When implemented in software, the above-described embodiments can be implemented in the form of one or more computer programs that are stored in a computer-readable storage medium. The computer-readable storage medium stores one or more computer instructions or computer programs that, when loaded into a computer, cause the computer to perform the processes or functions described in the embodiments of the present application. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable apparatus. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another computer-readable storage medium, such as from a website, a computer, a server, or a data center to another website, computer, server, or data center, via a wired (e.g., infrared, wireless, microwave, etc.) manner. The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server, data center, etc. that contains one or more collections of available media. The available media can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium. The semiconductor medium can be a solid-state disk.

[0172] It should be understood that the size of the sequence number of each process described above in various embodiments of the present application does not mean the order of execution, and the execution order of each process should be determined by its function and inherent logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0173] Those of ordinary skill in the art can realize that the units and algorithm steps of the examples described in conjunction with the embodiments disclosed herein can be realized in electronic hardware or a combination of computer software and electronic hardware. Whether the functions are realized in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the present application.

[0174] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the devices, apparatuses and units described above can refer to the corresponding processes in the foregoing method embodiments, which will not be described here.

[0175] In several embodiments provided by the present application, it should be understood that the disclosed devices, apparatuses and methods can be implemented in other manners. For example, the embodiments of the apparatus described above are merely schematic. For example, the division of the units is only a logical function division. There can be another division manner for the actual implementation. For example, a plurality of units or components can be combined or integrated into another device, or some features can be ignored or not executed. In addition, the displayed or discussed mutual couplings or direct couplings or communication connections can be indirect couplings or communication connections through some interfaces, devices or units, and can be in electrical, mechanical or other forms.

[0176] The units described as separate components can or can not be physically separate, and the components shown as units can or can not be physical units, i.e., can be located in one place, or can be distributed on a plurality of network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the embodiment.

[0177] In addition, each functional unit in each embodiment of the present application can be integrated into a processing unit, or each unit can be physically present separately, or two or more units can be integrated into one unit.

[0178] If the functions are realized in the form of software function units and sold or used as independent products, they can be stored in a computer readable storage medium. Based on this understanding, the technical solutions of the present application or the part of the present application that essentially contributes to the prior art or the part of the technical solutions can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes a plurality of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and various media that can store program codes.

[0179] The embodiment of the present application provides a readable storage medium, which includes: a program or instructions stored on the readable storage medium, the program or instructions are executed by a processor to realize the steps of the dorsalis pedis artery blood flow state monitoring method described above, and the same technical effect can be achieved. To avoid repetition, the present application will not be described again.

[0180] It should be pointed out finally that the above embodiments are only used to illustrate the technical solutions of the embodiments of the present application, but not to limit them. Although the present application has been described in detail with reference to the foregoing embodiments, it should be understood by those skilled in the art that the technical solutions recorded in the foregoing embodiments can be modified, or some technical features can be replaced equivalently; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application. Any changes or replacements that can be easily thought of by those skilled in the art within the technical scope disclosed by the present application should be covered within the protection scope of the present application.

Claims

1. A method of monitoring the state of blood flow in the dorsalis pedis artery, characterized by, The method comprises the following steps: S1: Collecting multi-source optical signals of the dorsal artery of the foot by an integrated probe; S2: Extracting morphological features of the pre-processed multi-source optical signals by a pulse wave analysis method; S3: Mapping the morphological features by combining a pre-trained multiple linear regression model to determine the local blood pressure; S4: According to the local blood pressure, combining the principle of vascular resistance, evaluating the stiffness of the peripheral vessels and the microvascular diastolic function in the dorsal artery region of the foot; S5: According to the evaluation results, constructing the transfer function between the local blood pressure and the blood flow velocity in the multi-source optical signals, and determining the dorsal region automatic regulation index through the transfer function; S6: Fusing the morphological features, the local blood pressure, the dorsal region automatic regulation index and the static risk factors of the patient's electronic medical record by a light gradient boosting tree model to determine the dorsal artery comprehensive perfusion index; S7: According to the dorsal artery comprehensive perfusion index, designing a context-aware adaptive alarm logic for the foot dorsal artery blood flow state monitoring after femoral artery intervention; S8: According to the context-aware adaptive alarm logic, outputting a visual report of the foot dorsal artery blood flow state monitoring result.

2. The dorsalis pedis artery blood flow state monitoring method according to claim 1, characterized by, Wherein, The multi-source optical signals include the blood flow velocity of the main trunk of the dorsal artery of the foot, the microvascular pulse waveform and the microcirculation blood flow of the dorsal region.

3. The dorsalis pedis artery blood flow state monitoring method according to claim 1, characterized by, The S2 specifically comprises: S201: According to the trough position of the microvascular pulse waveform in the pre-processed multi-source optical signals, the microvascular pulse waveform is divided into multiple single cardiac cycle pulses; S202: Within the monitoring time window, constructing the epoch pulse waveform for each single cardiac cycle pulse; S203: Extracting the key feature points of the epoch pulse waveform and the time stamps corresponding to the key feature points by the pulse wave analysis method; S204: Constructing the morphological features according to the key feature points and the time stamps corresponding to the key feature points.

4. The dorsalis pedis artery blood flow state monitoring method according to claim 1, characterized by, The training process of the pre-trained multiple linear regression model in S3 specifically comprises: S301: Combining the morphological features and the reference local blood pressure, a training sample set is constructed, wherein the training sample set includes multiple groups of records and foot epoch samples, the multiple groups of records include multiple foot epoch samples, and the foot epoch sample includes a morphological feature vector and a corresponding reference local blood pressure label; S302: Inputting the training sample set into the multiple linear regression model to output an initial parameter vector; S303: According to the initial parameter vector and the reference local blood pressure, calculating the individualized offset of the systolic pressure and the diastolic pressure; S304: Jointly compensating the individualized offset by using multiple groups of morphological features and their corresponding reference local blood pressure data for model parameter updating until the estimation deviation of the multiple linear regression model on each group of records remains stable to determine the final parameter vector of the multiple linear regression model; S305: According to the final parameter vector, the pre-trained multiple linear regression model is determined.

5. The dorsalis pedis artery blood flow state monitoring method according to claim 1, characterized by, The S4 specifically comprises: S401: calculating a local mean arterial pressure of the dorsal artery region according to the local blood pressure; S402: calculating a peripheral vascular resistance according to the local mean arterial pressure and a microcirculation blood flow of the dorsal region based on the vascular resistance principle; S403: evaluating a stiffness of the peripheral blood vessel and a microvascular dilation function according to a change of the peripheral vascular resistance.

6. The dorsalis pedis artery blood flow state monitoring method according to claim 5, wherein, The S5 specifically includes: S501: constructing an input-output signal sequence according to the local mean arterial pressure and the blood flow velocity; S502: performing Fourier transform on the input-output signal sequence; S503: calculating a transfer function between the local blood pressure and the blood flow velocity in the multi-source optical signal based on the Fourier transform result; S504: calculating gain and phase characteristics of the transfer function in a specific frequency band; S505: calculating the dorsal region automatic regulation index according to the gain and the phase characteristics in combination with a rate index.

7. The dorsalis pedis artery blood flow state monitoring method according to claim 1, characterized by, The S6 specifically includes: S601: constructing an epoch-level fusion feature input vector based on the morphological characteristics, the local blood pressure, the dorsal region automatic regulation index, and static risk factors of the patient's electronic medical record; S602: merging feature subsets with mutual exclusion in the epoch-level fusion feature input vector through an EFB mechanism to generate a bound feature; S603: calculating a bin value of the bound feature; S604: constructing a histogram statistic according to the bound feature and the bin value corresponding to the bound feature; S605: calculating a plurality of variance gain values corresponding to candidate split points according to the histogram statistic; S606: introducing a sample weight correction mechanism based on gradient size to correct the variance gain values of the candidate split points; S607: determining an optimal discriminant output of the bound feature according to the corrected variance gain values; S608: performing forward discriminant calculation on the current epoch-level fusion feature input vector through the lightweight gradient boosting tree model based on the optimal discriminant output to determine the dorsal artery comprehensive perfusion index.

8. The dorsalis pedis artery blood flow state monitoring method according to claim 1, characterized by, The S7 specifically includes: S701: constructing a postoperative monitoring situation state vector in combination with a postoperative time factor and the dorsal artery comprehensive perfusion index; S702: setting different reference perfusion threshold values for different postoperative stages according to the postoperative monitoring situation state vector; S703: calculating a change rate of the dorsal artery comprehensive perfusion index in a sliding time window according to the reference perfusion threshold value; S704: determining a normalized trend indicator of the dorsal artery comprehensive perfusion index according to the change rate; S705: constructing a multi-condition risk discriminant function according to the normalized trend indicator in combination with the dorsal region automatic regulation index; S706: designing the situation-aware adaptive alarm logic based on the multi-condition risk discriminant function.

9. The dorsalis pedis artery blood flow state monitoring method according to claim 1, characterized by, The dorsal artery blood flow state monitoring result report specifically includes: A dorsalis pedis arterial composite perfusion index and its trend over time, local blood pressure parameters and hemodynamic parameters corresponding to said composite perfusion index, an automatic regulation ability assessment result of the dorsalis pedis region, an alarm level generated based on a context-aware adaptive alarm logic and its triggering basis, and reference threshold information and individualized reference information corresponding to post-femoral arterial intervention phases.

10. A dorsalis pedis arterial blood flow state monitoring system characterized by, Comprises: a processor and a memory; said memory stores programs or instructions executable on said processor, said programs or instructions, when executed by said processor, implement the steps of the dorsalis pedis arterial blood flow state monitoring method as claimed in any one of claims 1 to 9.

Images