Search

CN-122004979-A - Pressure feedback-based self-adaptive control method and system for intelligent hemostatic device

CN122004979ACN 122004979 ACN122004979 ACN 122004979ACN-122004979-A

Abstract

The invention provides an intelligent hemostatic device self-adaptive control method and system based on pressure feedback, and relates to the technical field of self-adaptive control. And (3) screening an optimal target scheme through virtual previewing simulation, driving the pressing unit to execute, and updating model parameters in real time based on execution feedback. The method realizes the accurate improvement of hemostatic efficacy and the effective reduction of the risk of tissue injury.

Inventors

  • BAI XUE

Assignees

  • 中国人民解放军总医院第一医学中心

Dates

Publication Date
20260512
Application Date
20260311

Claims (10)

  1. 1. The self-adaptive control method of the intelligent hemostatic device based on pressure feedback is characterized by comprising the following steps of: collecting real-time pressure distribution data and physiological state characteristic parameters of a contact interface between the hemostatic device and a wound part; constructing a pressure-physiological coupling response model based on the real-time pressure distribution data and the physiological state characteristic parameters, establishing a bidirectional prediction function of pressure regulation and control behaviors on hemostatic efficacy and tissue injury by extracting dynamic correlation characteristics between pressure change rate and physiological response delay, and outputting a safe pressure application interval in the current physiological state; performing differential driving distribution on a plurality of pressing units of the hemostatic device according to the safe pressure application interval, and generating a partition cooperative initial pressure application scheme by taking upper and lower boundary values of the safe pressure application interval as constraint conditions of each pressing unit; Performing virtual previewing simulation on the initial pressure application scheme, calculating hemostatic timeliness scores and safety margin scores of all candidate schemes by simulating a pressure conduction path and a physiological response chain evolution process, and screening out a target pressure application scheme with double scores weighted and comprehensive optimal; And driving each pressure applying unit to execute pressure regulation according to the target pressure applying scheme, collecting the pressure deviation amount and the physiological state drift amount after execution in real time, transmitting the pressure deviation amount and the physiological state drift amount back to the pressure-physiological coupling response model, and updating the parameters of the dynamic association characteristic and the bidirectional prediction function.
  2. 2. The method of claim 1, wherein constructing a pressure-physiological coupling response model based on the real-time pressure distribution data and the physiological state characteristic parameters, establishing a bi-directional prediction function of pressure regulation behavior on hemostatic efficacy and tissue injury by extracting dynamic correlation characteristics between pressure change rate and physiological response delay, and outputting a safe pressure application interval in a current physiological state comprises: Performing time sequence differential processing on the real-time pressure distribution data to obtain a pressure change rate time sequence, and synchronously extracting physiological response state transition time marks in physiological state characteristic parameters; Performing time axis association analysis on the pressure change rate time sequence and the physiological response state transition time mark, calculating the time interval between the pressure application peak time and the physiological response state transition time, and extracting dynamic association features representing the pressure-physiological conduction delay law; constructing a pressure-physiological coupling response model based on the dynamic correlation characteristics, taking the time interval in the dynamic correlation characteristics as a time lag parameter of the model, and establishing a bidirectional prediction function; And calculating a minimum pressure threshold required for reaching preset hemostatic efficacy through a positive predictive function branch of the bidirectional predictive function, calculating a maximum pressure threshold corresponding to triggering tissue injury risk through a negative predictive function branch of the bidirectional predictive function, and determining a section between the minimum pressure threshold and the maximum pressure threshold as a safe pressure application section.
  3. 3. The method of claim 2, wherein performing a time-axis correlation analysis of the pressure change rate time series and the physiological response state transition time stamp, calculating a time interval between a pressure application peak time and a physiological response state transition time, and extracting a dynamic correlation feature characterizing a pressure-physiological conduction delay law comprises: Performing peak identification processing on the pressure change rate time sequence, detecting the moment when the pressure change rate in the pressure change rate time sequence reaches a local maximum value, and determining the detected moment as the pressure application peak moment; Time sequence alignment is carried out on the pressure application peak time and the physiological response state transition time mark on a unified time axis, and for the pressure application peak time, a physiological response state transition time mark with a time stamp positioned behind the pressure application peak time and closest to the time is searched in the physiological response state transition time mark, so that a causal pairing relation between the pressure application peak time and the physiological response state transition time mark is established; calculating a time difference value between the pressure application peak time and the paired physiological response state transition time mark based on the causal pairing relation, taking the time difference value as a time interval of single pressure-physiological conduction delay, and collecting a plurality of groups of time intervals corresponding to the causal pairing relation to form a time interval data set; And carrying out statistical analysis on the time interval data set, calculating a central tendency metric value and a discrete degree metric value of the time interval data set, and combining the central tendency metric value and the discrete degree metric value to form a dynamic correlation characteristic.
  4. 4. The method of claim 1, wherein differentially driving the plurality of pressure applying units of the hemostatic device according to the safe pressure applying interval, wherein generating the zoned coordinated initial pressure applying scheme using upper and lower boundary values of the safe pressure applying interval as constraints of each pressure applying unit comprises: obtaining hemostasis demand measurement values of wound contact areas corresponding to a plurality of pressing units of the hemostasis device, and carrying out urgency sequencing on the pressing units based on the hemostasis demand measurement values to obtain urgency serial numbers of the pressing units; According to the urgency sequence number, linear mapping distribution is carried out in the safety pressure application interval, a pressure unit with the forefront urgency sequence number is mapped to an upper boundary value of the safety pressure application interval, a pressure unit with the last urgency sequence number is mapped to a lower boundary value of the safety pressure application interval, equal proportion interpolation is carried out on other pressure units between the upper boundary value and the lower boundary value according to the urgency sequence number, and an initial pressure distribution value of each pressure unit is formed; Taking the initial pressure distribution value as a pressure reference value of each pressure applying unit, acquiring spatial position information of each pressure applying unit, and calculating a difference value between the spatial distance between adjacent pressure applying units and the pressure reference value; Constructing pressure conduction influence coefficients among the pressure applying units based on the difference value between the space distance and the pressure reference value, and carrying out neighborhood collaborative correction on the pressure reference value of each pressure applying unit by utilizing the pressure conduction influence coefficients to obtain collaborative correction pressure values of each pressure applying unit; And carrying out boundary constraint verification on the collaborative correction pressure value and the upper and lower boundary values of the safety pressure application interval, taking the verified collaborative correction pressure value as a target execution pressure value of each pressure application unit, and combining the target execution pressure value and a corresponding pressure application unit mark to form an initial pressure application scheme.
  5. 5. The method of claim 4, wherein constructing a pressure conduction influence coefficient between the pressure applying units based on the difference between the spatial distance and the pressure reference value, and performing neighborhood collaborative correction on the pressure reference value of each pressure applying unit by using the pressure conduction influence coefficient, and obtaining collaborative corrected pressure values of each pressure applying unit comprises: Aiming at a target pressing unit in a plurality of pressing units, a neighborhood pressing unit which is adjacent to the target pressing unit in space is obtained, a space distance value between the target pressing unit and each neighborhood pressing unit is calculated, and meanwhile, a pressure reference value difference value between the target pressing unit and each neighborhood pressing unit is calculated; Performing coupling operation on the space distance value and the pressure reference value difference value, and performing product calculation on the pressure reference value difference value and the reciprocal of the space distance value to obtain pressure conduction influence coefficients of each neighborhood pressure applying unit on the target pressure applying unit; Normalizing the pressure conduction influence coefficients to ensure that the sum of the pressure conduction influence coefficients corresponding to all the neighborhood pressure applying units is a preset unit value, and obtaining normalized pressure conduction influence coefficients; Taking the normalized pressure conduction influence coefficient as a weight coefficient, carrying out weighted summation on pressure reference values of all neighborhood pressure applying units to obtain a neighborhood pressure weighted average value, and calculating a deviation value between the neighborhood pressure weighted average value and the pressure reference value of the target pressure applying unit; And scaling the deviation value according to a preset correction proportion coefficient to obtain a pressure correction increment, and superposing the pressure reference value of the target pressing unit and the pressure correction increment to obtain a cooperative correction pressure value of the target pressing unit.
  6. 6. The method of claim 1, wherein performing a virtual pre-modeling of the initial pressure application protocol, calculating a hemostatic timeliness score and a safety margin score for each candidate protocol by modeling a pressure conduction path and a physiological response chain evolution process, and screening out a target pressure application protocol with a dual score weighted composite optimum comprises: obtaining target execution pressure values of each pressure applying unit in the initial pressure applying scheme, and constructing a simulation environment model comprising the elastic modulus distribution of the wound tissue and the topological structure of the vascular network; Calculating a stress tensor propagation process of pressure based on tissue elastic modulus distribution in the simulation environment model, tracking a diffusion track of the pressure from each pressurizing unit to deep tissues, and generating pressure field evolution data comprising a conduction time sequence and a spatial distribution gradient; Performing spatial mapping on the pressure field evolution data and the vascular network topological structure, calculating the vascular closure degree according to pressure values at each vascular node, and constructing a blood flow blocking dynamic process based on the time variation trend of the vascular closure degree to obtain a physiological response chained evolution sequence; Positioning critical time when all blood vessel nodes reach a complete closure state by analyzing a time evolution curve of a blood vessel node closure state in the physiological response chain evolution sequence, calculating time deviation between a preset hemostasis time target value and the critical time as a hemostasis timeliness score, capturing a stress peak break point by monitoring a stress accumulation process of a tissue unit in the physiological response chain evolution sequence, and calculating safety redundancy of a preset safety stress threshold and the stress peak break point as a safety margin score; And respectively endowing the hemostatic timeliness score and the safety margin score with preset weight coefficients, summing to obtain a comprehensive score, executing a simulation flow on a plurality of candidate initial pressure applying schemes, and selecting a candidate scheme with the largest comprehensive score as a target pressure applying scheme.
  7. 7. The method of claim 6, wherein calculating a stress tensor propagation process of pressure based on the tissue elastic modulus distribution in the simulated environmental model, tracking a diffusion trajectory of pressure from each pressurizing unit to the deep tissue, generating pressure field evolution data comprising a conduction time sequence and a spatial distribution gradient, comprises: acquiring the space grid division data of the wound tissue in the simulation environment model, and mapping the target execution pressure value of each pressure applying unit to a corresponding grid unit to serve as an initial stress state; Establishing a stress balance equation between grid cells based on the initial stress state and the tissue elastic modulus value of each grid cell, calculating a stress transfer relation between adjacent grid cells by solving the stress balance equation, and iteratively calculating a stress tensor evolution value of each grid cell under continuous time steps to form a stress tensor space-time evolution matrix; Extracting pressure components of each grid unit under each time step from the stress tensor space-time evolution matrix, carrying out layering statistics on the pressure components according to the depth coordinates of the grid units, calculating the attenuation proportion of each layering of the pressure from the surface layer where the pressure applying unit is positioned to the deep tissue, and generating a pressure depth penetration curve; correlating the pressure value of each depth layer in the pressure depth penetration curve with the transverse space coordinates of grid units in the corresponding layer, calculating the pressure difference distribution of different transverse positions in the same depth layer, and constructing a two-dimensional pressure diffusion track comprising a depth dimension and a transverse dimension; Calculating the pressure change rate between adjacent grid units in the two-dimensional pressure diffusion track, taking the pressure change rate as a spatial distribution gradient value, binding the spatial distribution gradient value with time step information in the stress tensor space-time evolution matrix, and generating pressure field evolution data comprising a conduction time sequence and a spatial distribution gradient.
  8. 8. An intelligent hemostatic device adaptive control system based on pressure feedback for implementing the method of any of the preceding claims 1-7, comprising: The data acquisition unit is used for acquiring real-time pressure distribution data and physiological state characteristic parameters of a contact interface between the hemostatic device and the wound part; The model construction unit is used for constructing a pressure-physiological coupling response model based on the real-time pressure distribution data and the physiological state characteristic parameters, establishing a bidirectional prediction function of pressure regulation and control behaviors on hemostatic efficacy and tissue injury by extracting dynamic correlation characteristics between pressure change rate and physiological response delay, and outputting a safe pressure application interval in the current physiological state; the scheme generating unit is used for carrying out differential driving distribution on a plurality of pressing units of the hemostatic device according to the safe pressure applying interval, and generating a partition cooperative initial pressure applying scheme by taking upper and lower boundary values of the safe pressure applying interval as constraint conditions of the pressing units; The simulation optimizing unit is used for carrying out virtual previewing simulation on the initial pressure applying scheme, calculating hemostatic timeliness scores and safety margin scores of all candidate schemes by simulating a pressure conduction path and a physiological response chained evolution process, and screening out a target pressure applying scheme with double scores weighted comprehensive optimization; And the feedback updating unit is used for driving each pressure applying unit to execute pressure regulation and control according to the target pressure applying scheme, collecting the pressure deviation and the physiological state drift after execution in real time, transmitting the pressure deviation and the physiological state drift back to the pressure-physiological coupling response model, and updating the parameters of the dynamic association characteristic and the bidirectional prediction function.
  9. 9. An electronic device, comprising: A processor; A memory for storing processor-executable instructions; Wherein the processor is configured to invoke the instructions stored in the memory to perform the method of any of claims 1 to 7.
  10. 10. A computer readable storage medium having stored thereon computer program instructions, which when executed by a processor, implement the method of any of claims 1 to 7.

Description

Pressure feedback-based self-adaptive control method and system for intelligent hemostatic device Technical Field The invention relates to an adaptive control technology, in particular to an adaptive control method and an adaptive control system for an intelligent hemostatic device based on pressure feedback. Background In the field of wound emergency and surgical operations, effective control of bleeding is a critical link in life saving. Traditional hemostatic methods typically rely on manual compression, tourniquets, or fixed pressure hemostatic devices. These methods generally employ a static or empirical pressure application pattern, i.e., an operator applies a predetermined or empirically determined constant pressure, depending on the type and location of the wound. For tourniquets and the like, the pressure settings are often based on population average data and remain fixed during use. The core logic of such methods is to close the vessel end by sustained external mechanical compression, which is relatively simple to implement and has long been demonstrated by clinical practice. However, the conventional practice has the obvious limitation that the static pressure mode cannot adapt to the dynamically changing physiological state of the wound site. The hemodynamic state, extent of tissue edema, and local metabolic environment of human tissue at various points in time are constantly changing, with a fixed pressure being insufficient to achieve effective hemostasis at some points in time and excessive pressure causing ischemia, nerve damage, or pressure necrosis of the distal tissue at other points in time. The prior art lacks an evaluation and feedback mechanism for real-time, quantitative correlation between pressure-exerted effects and physiological states. It is difficult for the operator to know precisely whether the currently applied pressure is within a "safety window" that is both effective in stopping bleeding and minimizing tissue damage, and further fine adjustments are not possible based on individual differences and real-time physiological responses. The unidirectional and open-loop control mode ensures that the hemostatic process has contradiction between safety and effectiveness, excessively depends on the experience of operators, and has insufficient accuracy and adaptability of overall control. Disclosure of Invention The embodiment of the invention provides an intelligent hemostatic device self-adaptive control method and system based on pressure feedback, which can solve the problems in the prior art. In a first aspect of an embodiment of the present invention, there is provided an adaptive control method for an intelligent hemostasis device based on pressure feedback, including: collecting real-time pressure distribution data and physiological state characteristic parameters of a contact interface between the hemostatic device and a wound part; constructing a pressure-physiological coupling response model based on the real-time pressure distribution data and the physiological state characteristic parameters, establishing a bidirectional prediction function of pressure regulation and control behaviors on hemostatic efficacy and tissue injury by extracting dynamic correlation characteristics between pressure change rate and physiological response delay, and outputting a safe pressure application interval in the current physiological state; performing differential driving distribution on a plurality of pressing units of the hemostatic device according to the safe pressure application interval, and generating a partition cooperative initial pressure application scheme by taking upper and lower boundary values of the safe pressure application interval as constraint conditions of each pressing unit; Performing virtual previewing simulation on the initial pressure application scheme, calculating hemostatic timeliness scores and safety margin scores of all candidate schemes by simulating a pressure conduction path and a physiological response chain evolution process, and screening out a target pressure application scheme with double scores weighted and comprehensive optimal; And driving each pressure applying unit to execute pressure regulation according to the target pressure applying scheme, collecting the pressure deviation amount and the physiological state drift amount after execution in real time, transmitting the pressure deviation amount and the physiological state drift amount back to the pressure-physiological coupling response model, and updating the parameters of the dynamic association characteristic and the bidirectional prediction function. Based on the real-time pressure distribution data and the physiological state characteristic parameters, constructing a pressure-physiological coupling response model, and by extracting dynamic correlation characteristics between a pressure change rate and physiological response delay, establishing a bidirectional prediction function of pr