Hysteroscopic surgery early-warning method, apparatus, and system, and storage medium

Abstract

A hysteroscopic surgery early-warning method, apparatus, and system, and a storage medium. The present invention relates to the technical field of surgery early-warning and is used for solving the problem in the prior art that hysteroscopic surgery early-warning relies on a single data source and has relatively poor practicability. Specifically, the method comprises: acquiring real-time electrocoagulation and electrosection data, the real-time negative deficit of uterine distension fluid, and the current surgery type (101); inputting the real-time electrocoagulation and electrosection data and the real-time negative deficit into a negative deficit prediction model to obtain a predicted negative deficit output by the negative deficit prediction model (102); obtaining, on the basis of a preset map, a first type of early-warning level and a first type of early-warning threshold range that correspond to the current surgery type (103); and using the first type of early-warning level corresponding to the threshold range within which the predicted negative deficit falls as a predicted early-warning level, and performing an early-warning operation corresponding to the predicted early-warning level (104). In this way, the present invention enables comprehensive utilization of a plurality of data sources in hysteroscopic surgery early warning; that is, the present invention can give an early warning of the surgery type while being capable of making predictions, showing improved practicability.

Description

Hysteroscopy operation early warning method, device, system and storage medium

[0001] Cross-reference to Related Applications

[0002] The present application claims priority to the Chinese patent application No. 2024112451960, filed on September 5, 2024, and entitled "Hysteroscopy operation early warning method, device and system", the content of which is incorporated herein by reference in its entirety. TECHNICAL FIELD

[0003] The present application relates to the technical field of operation early warning, in particular to a hysteroscopy operation early warning method, device, system and storage medium. BACKGROUND

[0004] Excessive absorption of uterine distension fluid during hysteroscopy operation can cause pulmonary edema, nervous system complications and even death, which is one of the serious complications of hysteroscopy operation. How to quickly and accurately calculate the inflow and outflow of uterine distension fluid has been a problem for clinical medical staff.

[0005] At present, the existing hysteroscopy operation early warning is achieved by monitoring and displaying negative value, uterine cavity pressure value and negative value change, and performing sound warning when the negative value or uterine cavity pressure value meets the requirements, so as to reduce the risk of uterine distension fluid over-absorption-hydroula overload syndrome. However, the data source monitored by the hysteroscopy operation early warning method is relatively single, and cannot fully reflect the complexity and potential risks of the operation process. Moreover, the existing hysteroscopy operation early warning only triggers the alarm by comparing the real-time detection data with the preset threshold, which lacks the prediction ability of future operation state changes and has poor practicability. Therefore, it is an urgent technical problem to provide a hysteroscopy operation early warning method with multiple data sources and practicability.

[0006] SUMMARY

[0007] Therefore, the purpose of the present application includes, for example, providing a hysteroscopy operation early warning method, device, system and storage medium to effectively solve the above technical problems.

[0008] The technical scheme provided by the embodiments of the present application is as follows:

[0009] In one aspect, the present application provides a hysteroscopy operation early warning method, comprising:

[0010] obtaining real-time electrocoagulation and electrocision data, real-time negative amount of uterine distension fluid and current operation type; wherein the real-time electrocoagulation and electrocision data are real-time electrocoagulation pedal pressure of hysteroscopy electrocoagulation control pedal and real-time electrocision pedal pressure of hysteroscopy electrocision control pedal; the real-time negative amount of uterine distension fluid is determined by real-time inflow weight and real-time outflow weight of uterine distension fluid;

[0011] inputting the real-time electrocoagulation and cutting data and the real-time negative balance of the uterine distending fluid into the negative balance prediction model to obtain a predicted negative balance output by the negative balance prediction model;

[0012] obtaining a first warning level corresponding to the current surgery type and a first warning threshold range corresponding to the first warning level based on a mapping relationship between the preset surgery type, the first warning level and the first warning threshold range;

[0013] taking the first warning level corresponding to the threshold range to which the predicted negative balance belongs as a predicted warning level, and performing a warning operation corresponding to the level of the predicted warning level.

[0014] Optionally, after obtaining the real-time electrocoagulation and cutting data, the real-time negative balance of the uterine distending fluid and the current surgery type, the method further comprises:

[0015] obtaining a historical negative balance of the uterine distending fluid, wherein the historical negative balance of the uterine distending fluid is determined by a historical inflow weight and a historical outflow weight of the uterine distending fluid;

[0016] determining a real-time change value of the negative balance based on the historical negative balance and the real-time negative balance;

[0017] determining whether the real-time change value is greater than a preset change threshold;

[0018] if yes, issuing a second type of operation risk prompt in a case where the real-time foot pedal pressure is not zero.

[0019] Optionally, after obtaining the real-time electrocoagulation and cutting data, the real-time negative balance of the uterine distending fluid and the current surgery type, the method further comprises:

[0020] obtaining case data of the patient and a type of the uterine distending fluid, wherein the case data comprises an age of the patient, a current condition of the patient and a basic condition of the patient;

[0021] inputting the age of the patient, the current condition of the patient and the type of the uterine distending fluid into a first threshold model to obtain a second warning level and a threshold range corresponding to the second warning level;

[0022] determining a target threshold range corresponding to the second warning level based on the basic condition of the patient and the threshold range corresponding to the second warning level;

[0023] determining a second warning level corresponding to the real-time negative balance based on the second warning level and the target threshold range corresponding to the second warning level, and taking the second warning level corresponding to the real-time negative balance as a second target warning level;

[0024] performing a warning operation corresponding to the second target warning level based on a corresponding relationship between the second warning level and the warning operation.

[0025] Optionally, based on the patient's underlying symptoms and the threshold range corresponding to the second-level warning, a target threshold range corresponding to the second-level warning is determined, including:

[0026] Determine whether the patient's underlying condition belongs to a pre-defined type of condition;

[0027] If so, the threshold range corresponding to each Category II warning level will be lowered, and the lowered threshold range will be used as the target threshold range corresponding to each Category II warning level.

[0028] If not, the threshold range corresponding to each second-class warning level output by the first threshold model evaluation model will be used as the target threshold range corresponding to each second-class warning level.

[0029] Optionally, hysteroscopic surgery early warning methods also include:

[0030] Obtain the patient's current intrauterine pressure and mean arterial pressure;

[0031] The difference between intrauterine pressure and mean arterial pressure is defined as the pressure difference, and the ratio of the pressure difference to the mean arterial pressure is defined as the pressure increase ratio.

[0032] Based on the preset third-class warning level and the threshold range corresponding to the third-class warning level, the third-class warning level corresponding to the boost ratio is determined, and the third-class warning level corresponding to the boost ratio is used as the third-class target warning level.

[0033] Based on the correspondence between the third-category warning level and the warning operation, the warning operation corresponding to the third-category target warning level shall be executed.

[0034] Optionally, the hysteroscopic surgery early warning method further includes:

[0035] After the surgery, save all electrocoagulation and electrosurgery data and negative volume data during the operation;

[0036] Based on all the electrocoagulation and electroresection data, behavioral data during the operation were determined, including the duration of each electroresection and electrocoagulation action, the total number of actions, and the total duration.

[0037] Establish the correlation between the current surgical type, negative volume data, and behavioral data, and save the correlation, as well as the current surgical type, distending fluid data, and behavioral data.

[0038] Optionally, before acquiring real-time electrocoagulation and resection data, the real-time negative / negative volume of distension fluid, and the current surgical type, the procedure may also include:

[0039] obtaining a training set; wherein the training set comprises a plurality of groups of monitoring data, wherein each group of monitoring data comprises electrocoagulation and electrocision data and negative underfill of uterine fluid at the same time;

[0040] inputting each group of monitoring data in the training set into the negative underfill prediction model respectively to obtain a negative underfill prediction result corresponding to each group of monitoring data; wherein the negative underfill prediction result comprises a predicted negative underfill at a first time after the collection time of the monitoring data;

[0041] determining a loss function value based on the negative underfill prediction result corresponding to each group of monitoring data and a negative underfill label result; wherein the negative underfill label result is the actual negative underfill at the first time after the collection time of the monitoring data;

[0042] updating the model parameters of the negative underfill prediction model based on the loss function value.

[0043] On the other hand, the application also provides a hysteroscopic surgery early warning device, comprising:

[0044] a data acquisition unit configured to obtain real-time electrocoagulation and electrocision data, real-time negative underfill of uterine distension fluid, and a current surgery type; wherein the real-time electrocoagulation and electrocision data are real-time electrocoagulation pedal pressure of a hysteroscopic electrocoagulation control pedal and real-time electrocision pedal pressure of a hysteroscopic electrocision control pedal; the real-time negative underfill of uterine distension fluid is determined by real-time inflow weight and real-time outflow weight of the uterine distension fluid;

[0045] a data prediction unit configured to input the real-time electrocoagulation and electrocision data and the real-time negative underfill into a negative underfill prediction model to obtain a predicted negative underfill output by the negative underfill prediction model;

[0046] a first threshold value determination unit configured to obtain a first early warning level corresponding to the current surgery type and a first early warning threshold value range corresponding to the first early warning level based on a mapping relationship between the preset surgery type, the first early warning level, and the first early warning threshold value range;

[0047] a first risk prompt unit configured to take the first early warning level corresponding to the threshold value range to which the predicted negative underfill belongs as a predicted early warning level, and issue a first operation risk prompt when the predicted early warning level is higher than a preset level.

[0048] Optionally, the hysteroscopic surgery early warning device further comprises:

[0049] a second risk prompt unit configured to obtain historical negative underfill of uterine distension fluid; wherein the historical negative underfill of uterine distension fluid is determined by historical inflow weight and historical outflow weight of the uterine distension fluid; determine a real-time change value of the negative underfill based on the historical negative underfill and the real-time negative underfill; judge whether the real-time change value is greater than a preset change threshold value; if yes, issue a second operation risk prompt in the case that the real-time pedal pressure is not zero.

[0050] Optionally, the hysteroscopic surgery early warning device further comprises:

[0051] The first early warning unit is configured to acquire case data of the patient and uterine distension liquid type; the case data includes patient age, current illness of the patient, and underlying illness of the patient; the patient age, the current illness of the patient, and the uterine distension liquid type are input into a first threshold model evaluation model to obtain a second type early warning level and a threshold range corresponding to the second type early warning level; based on the underlying illness of the patient and the threshold range corresponding to the second type early warning level, a target threshold range corresponding to the second type early warning level is determined; based on the second type early warning level and the target threshold range corresponding to the second type early warning level, a second type early warning level corresponding to the real-time negative volume is determined, and the second type early warning level corresponding to the real-time negative volume is taken as a second type target early warning level; based on a corresponding relationship between the second type early warning level and an early warning operation, an early warning operation corresponding to the second type target early warning level is performed.

[0052] Optionally, the first early warning unit is specifically configured to:

[0053] determine whether the underlying illness of the patient belongs to a preset type of illness;

[0054] if yes, the threshold range corresponding to each second type early warning level is lowered, and the lowered threshold range is taken as a target threshold range corresponding to each second type early warning level;

[0055] if no, the threshold range corresponding to each second type early warning level output by the first threshold model evaluation model is taken as a target threshold range corresponding to each second type early warning level.

[0056] Optionally, the hysteroscopic surgery early warning device further comprises:

[0057] The second early warning unit is configured to acquire the current intrauterine pressure and mean arterial pressure of the patient; determine a pressure difference between the intrauterine pressure and the mean arterial pressure, and determine a pressure increase ratio as a ratio of the pressure difference to the mean arterial pressure; based on a preset third type early warning level and a threshold range corresponding to the third type early warning level, a third type early warning level corresponding to the pressure increase ratio is determined, and the third type early warning level corresponding to the pressure increase ratio is taken as a third type target early warning level; based on a corresponding relationship between the third type early warning level and an early warning operation, an early warning operation corresponding to the third type target early warning level is performed.

[0058] Optionally, the hysteroscopic surgery early warning device further comprises:

[0059] The data storage processing unit is configured to save all the electrocoagulation and electrocision data and the negative volume data during the surgery after the surgery is completed; determine behavior data in the surgery based on all the electrocoagulation and electrocision data, wherein the behavior data includes single time length, total times and total time length of electrocoagulation and electrocision behavior; establish an association relationship among the current surgery type, the negative volume data and the behavior data, and save the association relationship and the current surgery type, the uterine distension liquid data and the behavior data.

[0060] Optionally, the hysteroscopic surgery early warning device further comprises:

[0061] The model training unit is configured to obtain a training set; wherein the training set includes a plurality of groups of monitoring data, wherein each group of monitoring data includes electrocoagulation and electrocision data and negative volume of uterine cavity liquid at the same time; input each group of monitoring data in the training set into the negative volume prediction model respectively to obtain a negative volume prediction result corresponding to each group of monitoring data; wherein the negative volume prediction result includes a predicted negative volume at a first time after the collection time of the monitoring data; determine a loss function value based on the negative volume prediction result corresponding to each group of monitoring data and a negative volume labeled result; wherein the negative volume labeled result is a true negative volume at the first time after the collection time of the monitoring data; update the model parameters of the negative volume prediction model based on the loss function value.

[0062] In another aspect, the embodiment of the present application provides a hysteroscopic surgery early warning system, comprising: a first weight sensor, a second weight sensor, a first pressure sensor, a second pressure sensor and a processor;

[0063] The first weight sensor is arranged at the bottom of the uterine distension liquid supply device, and the first weight sensor is configured to collect the inflow weight of the uterine distension liquid;

[0064] The second weight sensor is arranged at the bottom of the uterine distension liquid outflow storage device, and the second weight sensor is configured to collect the outflow weight of the uterine distension liquid;

[0065] The first pressure sensor is arranged at the bottom of the hysteroscopic electrocoagulation control pedal, and the first pressure sensor is configured to collect the electrocoagulation pedal pressure;

[0066] The second pressure sensor is arranged at the bottom of the hysteroscopic electrocision control pedal, and the second pressure sensor is configured to collect the electrocision pedal pressure;

[0067] The processor is connected with the first weight sensor, the second weight sensor, the first pressure sensor and the second pressure sensor respectively; and the processor is configured to execute the computer program corresponding to the hysteroscopic surgery early warning method provided by the embodiment of the present application.

[0068] In a possible implementation, the hysteroscopic surgery early warning system further comprises: a first flow sensor arranged on the inflow pipeline of the uterine distention liquid and a second flow sensor arranged on the outflow pipeline of the uterine distention liquid.

[0069] The first flow sensor is configured to collect the inflow flow of the uterine distention liquid, and the second flow sensor is configured to collect the outflow flow of the uterine distention liquid.

[0070] The hysteroscopic surgery early warning method has the following beneficial effects:

[0071] In the embodiment, the real-time electrocoagulation and electrocision data and the real-time negative volume are input into the negative volume prediction model to obtain the predicted negative volume output by the negative volume prediction model. The negative volume is predicted based on the electrocoagulation and electrocision data corresponding to the electrocoagulation and electrocision operation of the doctor during the surgery and the real-time negative volume, the comprehensive utilization of multiple data sources in the hysteroscopic surgery early warning process is realized, the negative volume prediction model constructed based on the multiple data sources can comprehensively and accurately predict the negative volume, and a good data basis is provided for the hysteroscopic surgery early warning. The prediction of the negative volume based on the predicted negative volume enables the hysteroscopic surgery early warning method to have the prediction ability for the future change of the negative volume, and the practicability is improved. The predicted early warning level is further determined based on the early warning level and the threshold value corresponding to the current surgery type, the early warning operation corresponding to the level of the predicted early warning level is performed, and the negative volume early warning for the current surgery type is realized, and the early warning effect of the hysteroscopic surgery early warning is improved.

[0072] Other features and advantages of the present application will be set forth in the following description, and in part will become apparent to those skilled in the art from the description, or can be learned by practice of the present application. The objects and other advantages of the present application will be realized and attained by the structure particularly pointed out in the written description and claims hereof as well as the appended drawings. BRIEF DESCRIPTION OF DRAWINGS

[0073] The accompanying drawings, which are included to provide a further understanding of the present application and are incorporated in and constitute a part of this application, illustrate embodiments of the present application and together with the description serve to explain the present application. In the drawings:

[0074] FIG. 1 is a schematic diagram of a general flow of a hysteroscopic surgery early warning method according to an embodiment of the present application;

[0075] FIG. 2 is a schematic diagram of a general flow of a real-time early warning method for a negative volume of uterine distention liquid in a hysteroscopic surgery according to an embodiment of the present application;

[0076] FIG. 3 is a schematic diagram of a general flow of a real-time early warning method for intrauterine pressure in a hysteroscopic surgery according to an embodiment of the present application;

[0077] FIG. 4 is a schematic diagram of a functional structure of a hysteroscopic surgery early warning device according to an embodiment of the present application;

[0078] Fig. 5 is a schematic diagram of the framework of the hysteroscopic surgery early warning system in the embodiments of the present application.

[0079] Icons: 500-hysteroscopic surgery early warning system; 501-first weight sensor; 502-second weight sensor; 503-first pressure sensor; 504-second pressure sensor; 505-processor; 506-first flow sensor 507-second flow sensor. DETAILED DESCRIPTION

[0080] In order to make the purposes, technical solutions and beneficial effects of the present application clearer, the technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only some of the embodiments of the present application, not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0081] The hysteroscopic surgery early warning method provided in the embodiments of the present application is described below with reference to Fig. 1. The general flow of the hysteroscopic surgery early warning method provided in the embodiments of the present application is as follows:

[0082] Step 101: acquiring real-time electrocoagulation and electrocision data, real-time negative amount of uterine distending fluid, and current surgery type; wherein the real-time electrocoagulation and electrocision data are real-time electrocoagulation pedal pressure of a hysteroscopic electrocoagulation control pedal and real-time electrocision pedal pressure of a hysteroscopic electrocision control pedal; the real-time negative amount of uterine distending fluid is determined by real-time inflow weight and real-time outflow weight of the uterine distending fluid.

[0083] In practical applications, surgical procedures include: hysteroscopic electrocautery, intrauterine adhesion lysis, myomectomy, hysteroscopic tubal patency testing, hysteroscopic endometriosis treatment, hysteroscopic embryo removal, and hysteroscopic foreign body removal. The current surgical procedure is at least one of these. A first pressure sensor is installed at the bottom of the hysteroscopic electrocoagulation control foot pedal, and the pressure collected in real time by the first pressure sensor is used as the real-time electrocoagulation foot pedal pressure. A second pressure sensor is installed at the bottom of the hysteroscopic electroresection control foot pedal, and the pressure collected in real time by the second pressure sensor is used as the real-time electroresection foot pedal pressure. Since a hysteroscope can only operate in one of the electroresection mode or the electrocoagulation mode, at any given time, at least one of the real-time electrocoagulation foot pedal pressure and the real-time electroresection foot pedal pressure will be zero. The distending fluid infusion device is equipped with a first weight sensor, and the distending fluid collection device is equipped with a second weight sensor. The weight collected in real time by the first weight sensor is the real-time infusion weight of the distending fluid, and the weight collected in real time by the second weight sensor is the real-time outflow weight of the distending fluid. The real-time deficit weight of the distending fluid at this time is obtained by subtracting the real-time outflow weight from the real-time infusion weight. The real-time deficit amount of the distending fluid is determined based on the real-time deficit weight and density of the distending fluid.

[0084] Step 102: Input the real-time electrocoagulation and electroswitching data and the real-time negative and negative quantities into the negative and negative quantity prediction model to obtain the predicted negative and negative quantities output by the negative and negative quantity prediction model.

[0085] In practical applications, the anti-collapse / overflow prediction model can be an observational model based on machine learning methods; specifically, it can employ a neural network model. Based on the input real-time electrocoagulation and electrocautery data and the real-time anti-collapse / overflow of the distending fluid, the anti-collapse / overflow prediction model can output a predicted anti-collapse / overflow at a future time. This "first time" is after a future time and separated from the future time by a preset time interval.

[0086] Step 103: Based on the preset mapping relationship between surgical type, first-class warning level and first-class warning threshold range, obtain the first-class warning level corresponding to the current surgical type and the first-class warning threshold range corresponding to the first-class warning level.

[0087] Step 104: Take the first type of early warning level corresponding to the threshold range to which the predicted negative amount belongs as the predicted early warning level, and execute the early warning operation corresponding to the level of the predicted early warning level.

[0088] In actual application, each type of operation corresponds to a first type of early warning level and a first type of early warning threshold range; each first type of early warning level is provided with a corresponding first type of early warning threshold range. The first type of early warning level is the level of early warning for the predicted negative volume, and the first type of early warning level can be divided into multiple levels, and the first type of early warning threshold range is the early warning threshold range for the predicted negative volume, which can include multiple non-intersecting negative volume ranges. Each level of the first type of early warning level corresponds to a negative volume range. As the level of the first type of early warning level increases, the negative volume range corresponding to the first type of early warning level increases, and the prediction risk of uterine distension fluid over-absorption gradually increases; each level of the first type of early warning level corresponds to different early warning operations. Specifically, the first type of early warning level can include a first type of first-level early warning, a first type of second-level early warning, and a first type of third-level early warning. The early warning operation corresponding to the first type of first-level early warning is to display the predicted early warning level and the predicted negative volume, and to light up the corresponding indicator light or make the indicator light display the corresponding color; the early warning operation corresponding to the first type of second-level early warning is to display the predicted early warning level and the predicted negative volume, light up the corresponding indicator light, and perform sound warning and issue a voice prompt corresponding to the risk of uterine distension fluid over-absorption, so as to enable the doctor to timely adjust the operation; the early warning operation corresponding to the first type of third-level early warning is to display the predicted early warning level and the predicted negative volume, light up the corresponding indicator light, and perform frequent sound warning and issue a voice prompt corresponding to a high risk of uterine distension fluid over-absorption, so as to enable the doctor to timely adjust the operation. In the case where there are many first type of early warning levels, multiple level thresholds can also be set, and the first type of early warning level in each level threshold range matches the same early warning operation, and the first type of early warning level in different level threshold ranges matches different early warning operations.

[0089] Optionally, after obtaining the electrocoagulation and electrocision data, the uterine distension fluid data, and the current operation type, the method further includes:

[0090] First, based on the historical negative volume and the real-time negative volume, a real-time change value of the negative volume is determined;

[0091] Then, it is judged whether the real-time change value is greater than a preset change threshold; if yes, an operation risk prompt is issued in the case where the real-time electrocoagulation pedal pressure or the real-time electrocision pedal pressure is not zero.

[0092] In actual application, the real-time change value of the negative volume is the real-time negative volume minus the historical negative volume, and the real-time change value of the negative volume greater than the preset change threshold corresponds to the rapid increase of the negative volume. The real-time electrocoagulation pedal pressure or the real-time electrocision pedal pressure not being zero means that the current hysteroscopy is in the electrocoagulation or electrocision operation. When it is determined that the real-time change value of the negative volume is greater than the preset change threshold and the real-time electrocoagulation pedal pressure or the real-time electrocision pedal pressure is not zero, it is determined that the electrocoagulation or electrocision operation causes the rapid increase of the negative volume, and an operation risk prompt is given, wherein the operation risk prompt corresponds to the case that the electrocoagulation or electrocision operation causes the rapid increase of the negative volume, and the operation risk prompt is at least one of the voice prompt, the corresponding indicator light, and the corresponding icon or text displayed on the display. When it is determined that the real-time change value of the negative volume is greater than the preset change threshold and the real-time electrocoagulation pedal pressure and the real-time electrocision pedal pressure are both zero, it is determined that other operations cause the rapid increase of the negative volume, and other risk prompts are given. The other risk prompt corresponds to the case that other operations cause the rapid increase of the negative volume, and the other risk prompt is at least one of the voice prompt, the corresponding indicator light, and the corresponding icon or text displayed on the display.

[0093] Optionally, referring to FIG. 2, after the electrocoagulation or electrocision data, the uterine distension liquid data, and the current operation type are obtained, the method further comprises:

[0094] Step 201: obtaining the case data of the patient and the uterine distension liquid type; wherein the case data includes the age of the patient, the current disease of the patient, and the basic disease of the patient.

[0095] In actual application, the uterine distension liquid type includes electrolyte solution and non-electrolyte solution. The current disease of the patient is the disease corresponding to the operation type. The current disease of the patient includes: endometrial polyp, submucous myoma, and abnormal endometrial hyperplasia corresponding to hysteroscopic resection; adhesion of uterine cavity corresponding to hysteroscopic adhesion separation; uterine myoma corresponding to hysteroscopic myomectomy; tubal obstruction corresponding to hysteroscopic tubal fluid infusion; endometriosis corresponding to hysteroscopic endometriosis treatment; early abortion, incomplete abortion, and missed abortion corresponding to hysteroscopic embryo removal; and intrauterine foreign body, embedded IUD, broken IUD, and ectopic IUD corresponding to hysteroscopic foreign body removal. The basic disease of the patient refers to the long-term disease of the patient.

[0096] Step 202: inputting the age of the patient, the current disease of the patient, and the uterine distension liquid type into the first threshold model to obtain the second type of early warning level and the threshold range corresponding to the second type of early warning level.

[0097] In actual application, each second type early warning level is provided with a corresponding second type early warning threshold range. The second type early warning level is an early warning level for real-time negative volume, the second type early warning level can be divided into multiple levels, and the second type early warning threshold range is an early warning threshold range for real-time negative volume, which can include multiple non-intersecting negative volume ranges. Each level in the second type early warning level corresponds to a negative volume range. As the level of the second type early warning level increases, the actual risk of current uterine distension fluid overabsorption increases.

[0098] In specific implementation, the first threshold evaluation model can output a corresponding second type early warning level and a threshold range corresponding to the second type early warning level according to the input patient age, patient current condition, and uterine distension fluid type. The first threshold evaluation model can be a neural network model or a linear regression model. The first threshold evaluation model is trained in the following manner.

[0099] First, a training set is obtained; wherein the training set includes multiple data groups, each data group including patient age and patient current condition, and the age and / or condition in each data group being different.

[0100] Then, each data group in the training set is input into the first threshold evaluation model to obtain a corresponding prediction result of each data group; wherein the prediction result includes a second type early warning level and a threshold range corresponding to the second type early warning level.

[0101] Finally, based on the prediction result and the target result corresponding to each data group, the model parameters of the first threshold evaluation model are updated; wherein the target result is the second type early warning level and the threshold range corresponding to the second type early warning level corresponding to the data group in the training set.

[0102] Step 203: determining a target threshold range corresponding to the second type early warning level based on the patient's underlying condition and the threshold range corresponding to the second type early warning level.

[0103] In actual application, since the body's bearing capacity is different when the patient has different underlying conditions, according to whether the patient's underlying condition belongs to a preset type of condition, it can be determined whether to adjust the threshold range corresponding to the second type early warning level, or the adjustment range of the threshold range corresponding to the second type early warning level. The preset type of condition includes comorbidities such as combined cardiovascular disease and renal failure. Different preset types of conditions can correspond to different preset adjustment ranges.

[0104] Specifically, based on the patient's underlying condition and the threshold range corresponding to the second type early warning level, the target threshold range corresponding to the second type early warning level can be determined in the following manner, but is not limited to the following manner:

[0105] First, it is determined whether the patient's underlying condition belongs to a preset type of condition;

[0106] Then, if the patient's underlying condition belongs to the preset type of condition, the threshold range corresponding to each second type of early warning level is down-regulated, and the down-regulated threshold range is taken as the target threshold range corresponding to each second type of early warning level; if the patient's underlying condition does not belong to the preset type of condition, the threshold range corresponding to each second type of early warning level output by the first threshold model is taken as the target threshold range corresponding to each second type of early warning level.

[0107] Step 204: Based on the second type of early warning level and the target threshold range corresponding to the second type of early warning level, the second type of early warning level corresponding to the real-time negative volume is determined, and the second type of early warning level corresponding to the real-time negative volume is taken as the second type of target early warning level.

[0108] Step 205: Based on the correspondence between the second type of early warning level and the early warning operation, the early warning operation corresponding to the second type of target early warning level is executed.

[0109] In actual application, the second type of early warning level is the early warning level of the real-time negative volume, and the second type of early warning level can include the second type of first-level early warning, the second type of second-level early warning, and the second type of third-level early warning. Each early warning level corresponds to a kind of early warning operation, the early warning operation corresponding to the second type of first-level early warning is to display the early warning level and the real-time negative volume, light up the corresponding indicator light or make the indicator light display the corresponding color; the early warning operation corresponding to the second type of first-level early warning is to display the early warning level and the real-time negative volume, light up the corresponding indicator light or make the indicator light display the corresponding color, and issue a voice prompt corresponding to the current uterine distension fluid absorption anomaly; the early warning operation corresponding to the second type of third-level early warning is to display the early warning level and the real-time negative volume, light up the corresponding indicator light or make the indicator light display the corresponding color, issue a voice prompt corresponding to the current uterine distension fluid over-absorption, and control the uterine distension fluid infusion device to stop infusing uterine distension fluid. With the increase of the second type of early warning level, the intensity and frequency of the alarm sound will increase.

[0110] Alternatively, since the systemic absorption of fluid significantly increases when the intrauterine pressure exceeds the mean arterial pressure. In order to reduce the risk of uterine distension fluid over-absorption, referring to FIG. 3, after obtaining the electrocoagulation and electrocision data, the uterine distension fluid data and the current type of operation, it further includes:

[0111] Step 301: Obtain the current intrauterine pressure and mean arterial pressure of the patient.

[0112] Step 302: Determine the difference between the intrauterine pressure and the mean arterial pressure as the pressure difference, and determine the ratio of the pressure difference to the mean arterial pressure as the pressure increase ratio.

[0113] Step 303: determining the third type of warning level corresponding to the boost ratio based on the preset third type of warning level and the threshold range corresponding to the third type of warning level, and taking the third type of warning level corresponding to the boost ratio as the third type of target warning level.

[0114] Step 304: performing the warning operation corresponding to the third type of target warning level based on the corresponding relationship between the third type of warning level and the warning operation.

[0115] In practical applications, the current intrauterine pressure is detected and sent by the uterine distension liquid supply device, and the average pulsatile pressure is collected and sent by the external blood pressure monitoring device. The third type of warning level includes the third type of first warning, the third type of second warning, and the third type of third warning. Correspondingly, the threshold range of the third type of first warning is 0-20%, the threshold range of the third type of second warning is 20-30%, and the threshold range of the third type of third warning is greater than 30%. Each warning level corresponds to an indicator light of a certain color. As the third type of warning level increases, the intensity and frequency of the alarm sound will increase.

[0116] Optionally, the hysteroscopic surgery warning method further comprises:

[0117] Firstly, after the surgery is completed, all the electrocoagulation and electrocision data and uterine distension liquid data during the surgery are saved;

[0118] Then, based on all the electrocoagulation and electrocision data and uterine distension liquid data, an electrocoagulation and electrocision data-uterine distension liquid data image is drawn and saved;

[0119] Next, based on all the electrocoagulation and electrocision data, behavior data during the surgery is determined, wherein the behavior data includes the single time length, total times, and total time length of the electrocision and electrocoagulation behavior;

[0120] Finally, an association relationship between the current surgery type, uterine distension liquid data, and behavior data is established, and the association relationship and the current surgery type, uterine distension liquid data, and behavior data are saved.

[0121] In practical applications, all the electrocoagulation and electrocision data and uterine distension fluid data during the operation can be configured to train the negative volume prediction model. Based on the electrocoagulation and electrocision data and the negative volume of the uterine distension fluid during the entire operation, the electrocoagulation and electrocision data-negative volume of the uterine distension fluid image is drawn with the electrocoagulation and electrocision data at the same time as the horizontal coordinate and the real-time negative volume of the uterine distension fluid as the vertical coordinate, wherein the image can include a plurality of segment data curves, and the data curve can be a curve between the electrocoagulation pedal pressure and the negative volume or a curve between the electrocision pedal pressure and the negative volume. During the operation of this type, the change relationship of the electrocoagulation and electrocision data-uterine distension fluid data image intuitively shows the influence of the electrocoagulation and electrocision behavior on the negative volume. A plurality of sets of correlation relationships, the current operation type, the uterine distension fluid data, and the behavior data are saved in the storage medium. For the same data of the current operation type, the uterine distension fluid data that meets the preset requirement can be filtered out, and the corresponding behavior data is obtained, the law of the behavior data is summarized, and the electrocoagulation and electrocision suggestion for the current operation type is given.

[0122] Optionally, the hysteroscopic operation early warning method further comprises:

[0123] When the operation duration exceeds the preset duration, the early warning of the operation is stopped.

[0124] Optionally, before the real-time electrocoagulation and electrocision data, the real-time negative volume of the uterine distension fluid, and the current operation type are obtained, the negative volume prediction model needs to be trained, and the specific training method can use but is not limited to the following method:

[0125] First, a training set is obtained; wherein the training set includes a plurality of sets of monitoring data, and each set of monitoring data includes electrocoagulation and electrocision data and negative volume of uterine cavity fluid at the same time;

[0126] Then, each set of monitoring data in the training set is input into the negative volume prediction model to obtain the negative volume prediction result corresponding to each set of monitoring data; wherein the negative volume prediction result includes the predicted negative volume at the first time after the collection time of the monitoring data;

[0127] Next, based on the negative volume prediction result corresponding to each set of monitoring data and the negative volume label result, a loss function value is determined; wherein the negative volume label result is the true negative volume at the first time after the collection time of the monitoring data;

[0128] Finally, based on the loss function value, the model parameters of the negative volume prediction model are updated.

[0129] In a specific implementation, each set of monitoring data in the training set can include the electrocoagulation and electrocision data and the negative underfill of uterine cavity fluid at the same time; the monitoring data in the training set can be obtained from the stored electrocoagulation and electrocision data and the uterine distension fluid data in the entire operation process. The loss function can be one of a mean square error function, a root mean square error function, and a mean absolute error function, and the loss function value output by the loss function can evaluate the difference between the negative underfill prediction result output by the negative underfill prediction model and the negative underfill expression result. The smaller the loss function value is, the higher the prediction accuracy of the negative underfill prediction model is. Based on the loss function value, it is determined whether the loss function value is less than a preset threshold value; the preset threshold value can be set according to actual needs. If the loss function value is not less than the preset threshold value, the model parameters of the negative underfill prediction model are updated, and retraining is performed until the loss function value is less than the preset threshold value, and the final negative underfill prediction model is obtained.

[0130] Based on the above embodiment, the hysteroscopic operation early warning device provided in the embodiment of the application comprises at least the following:

[0131] The data acquisition unit 401 is configured to acquire real-time electrocoagulation and electrocision data, real-time negative underfill of uterine distension fluid, and a current operation type; wherein the real-time electrocoagulation and electrocision data are real-time electrocoagulation pedal pressure of a hysteroscopic electrocoagulation control pedal and real-time electrocision pedal pressure of a hysteroscopic electrocision control pedal; the real-time negative underfill of uterine distension fluid is determined by real-time inflow weight and real-time outflow weight of the uterine distension fluid;

[0132] The data prediction unit 402 is configured to input the real-time electrocoagulation and electrocision data and the real-time negative underfill into a negative underfill prediction model to obtain a predicted negative underfill output by the negative underfill prediction model;

[0133] The first threshold value determination unit 403 is configured to obtain a first early warning level corresponding to the current operation type and a first early warning threshold value range corresponding to the first early warning level based on a mapping relationship between the preset operation type, the first early warning level, and the first early warning threshold value range;

[0134] The first risk prompt unit 404 is configured to take the first early warning level corresponding to the threshold value range to which the predicted negative underfill belongs as a predicted early warning level, and issue a first operation risk prompt when the predicted early warning level is higher than a preset level.

[0135] Optionally, the hysteroscopic operation early warning device further comprises:

[0136] The second risk prompting unit 405 is configured to acquire a historical negative amount of the uterine distension liquid; wherein the historical negative amount of the uterine distension liquid is determined by a historical filling weight and a historical outflow weight of the uterine distension liquid; a real-time change value of the negative amount is determined based on the historical negative amount and a real-time negative amount; it is judged whether the real-time change value is greater than a preset change threshold; if yes, a second type of operation risk prompt is issued in a case that the real-time foot pedal pressure is not zero.

[0137] Optionally, the hysteroscopic surgery early warning device further comprises:

[0138] The first early warning unit 406 is configured to acquire case data of the patient and a type of the uterine distension liquid; wherein the case data comprises a patient age, a current disease of the patient and a basic disease of the patient; the patient age, the current disease of the patient and the type of the uterine distension liquid are input into a first threshold model to obtain a second type of early warning level and a threshold range corresponding to the second type of early warning level; a target threshold range corresponding to the second type of early warning level is determined based on the basic disease of the patient and the threshold range corresponding to the second type of early warning level; a second type of early warning level corresponding to the real-time negative amount is determined based on the second type of early warning level and the target threshold range corresponding to the second type of early warning level, and the second type of early warning level corresponding to the real-time negative amount is taken as a second type of target early warning level; a warning operation corresponding to the second type of target early warning level is executed based on a corresponding relationship between the second type of early warning level and the warning operation.

[0139] Optionally, the first early warning unit 406 is specifically configured to:

[0140] determine whether the basic disease of the patient belongs to a preset type of disease;

[0141] if yes, the threshold range corresponding to each second type of early warning level is lowered, and the lowered threshold range is taken as a target threshold range corresponding to each second type of early warning level;

[0142] if no, the threshold range corresponding to each second type of early warning level output by the first threshold model is taken as a target threshold range corresponding to each second type of early warning level.

[0143] Optionally, the hysteroscopic surgery early warning device further comprises:

[0144] The second early warning unit 407 is configured to acquire a current intrauterine pressure and a mean arterial pressure of the patient; a difference between the intrauterine pressure and the mean arterial pressure is determined as a pressure difference, and a ratio of the pressure difference to the mean arterial pressure is determined as a pressurization ratio; a third type of early warning level corresponding to the pressurization ratio is determined based on a preset third type of early warning level and a threshold range corresponding to the third type of early warning level, and the third type of early warning level corresponding to the pressurization ratio is taken as a third type of target early warning level; a warning operation corresponding to the third type of target early warning level is executed based on a corresponding relationship between the third type of early warning level and the warning operation.

[0145] Optionally, the hysteroscopic surgery warning device further includes:

[0146] The data storage and processing unit 408 is configured to save all electrocoagulation and electroresection data and negative / negative volume data during the operation after the operation is completed; based on all the electrocoagulation and electroresection data, determine the behavioral data during the operation, wherein the behavioral data includes the single duration, total number of electroresection and electrocoagulation actions, and total duration; establish the correlation between the current operation type, negative / negative volume data, and behavioral data, and save the correlation, as well as the current operation type, distension fluid data, and behavioral data.

[0147] Optionally, the hysteroscopic surgery warning device further includes:

[0148] Model training unit 409 is configured to acquire a training set, which includes multiple sets of monitoring data. Each set of monitoring data includes electrocoagulation and electrocautery data and the negative / negative volume of intrauterine fluid at the same time. Each set of monitoring data in the training set is input into the negative / negative volume prediction model to obtain the negative / negative volume prediction result corresponding to each set of monitoring data. The negative / negative volume prediction result includes the predicted negative / negative volume at the first time after the data acquisition time. Based on the negative / negative volume prediction result and the negative / negative volume annotation result corresponding to each set of monitoring data, a loss function value is determined. The negative / negative volume annotation result is the actual negative / negative volume at the first time after the data acquisition time. Based on the loss function value, the model parameters of the negative / negative volume prediction model are updated.

[0149] It should be noted that the principle of the hysteroscopic surgery early warning device 400 provided in this application embodiment to solve the technical problem is similar to the hysteroscopic surgery early warning method provided in this application embodiment. Therefore, the implementation of the hysteroscopic surgery early warning device 400 provided in this application embodiment can refer to the implementation of the hysteroscopic surgery early warning method provided in this application embodiment, and the repeated parts will not be described again.

[0150] After introducing the hysteroscopic surgery early warning method and device provided in the embodiments of this application, the hysteroscopic surgery early warning system provided in the embodiments of this application will be briefly introduced next.

[0151] Referring to Figure 5, the hysteroscopic surgery early warning system 500 provided in this application embodiment includes at least: a first weight sensor 501, a second weight sensor 502, a first pressure sensor 503, a second pressure sensor 504, and a processor 505;

[0152] The first weight sensor 501 is located at the bottom of the distending fluid supply device, and the first weight sensor 501 is configured to collect the weight of the distending fluid injected.

[0153] The second weight sensor 502 is arranged at the bottom of the uterine distention fluid outflow storage device, and is configured to collect the outflow weight of the uterine distention fluid.

[0154] The first pressure sensor 503 is arranged at the bottom of the hysteroscope electrocoagulation control pedal, and is configured to collect the pressure of the electrocoagulation pedal.

[0155] The second pressure sensor 504 is arranged at the bottom of the hysteroscope electrocision control pedal, and is configured to collect the pressure of the electrocision pedal.

[0156] The processor is connected with the first weight sensor 501, the second weight sensor 502, the first pressure sensor 503 and the second pressure sensor 504 respectively, and is configured to execute the computer program corresponding to the hysteroscope surgery early warning method proposed in the embodiments of the present application.

[0157] Optionally, the hysteroscope surgery early warning system 500 further comprises a first flow sensor 506 arranged on the inflow pipeline of the uterine distention fluid and a second flow sensor 507 arranged on the outflow pipeline of the uterine distention fluid.

[0158] The first flow sensor 506 is configured to collect the inflow flow of the uterine distention fluid, and the second flow sensor 507 is configured to collect the outflow flow of the uterine distention fluid.

[0159] In a specific implementation, according to the inflow flow of the uterine distention fluid collected by the first flow sensor 506 and the outflow flow of the uterine distention fluid collected by the second flow sensor 507, a contrast negative balance of the uterine distention fluid can be further determined. The contrast negative balance can be used as backup data when the data collection of the first weight sensor 501 and / or the second weight sensor 502 is abnormal, that is, when the data collection of the first weight sensor 501 and / or the second weight sensor 502 is abnormal, the contrast negative balance determined according to the inflow flow of the uterine distention fluid collected by the first flow sensor 506 and the outflow flow of the uterine distention fluid collected by the second flow sensor 507 can be used as the real-time negative balance.

[0160] It should be noted that although several units or sub-units of the apparatus are mentioned in the foregoing detailed description, such a division is merely exemplary and not mandatory. Indeed, according to the embodiments of the present application, the features and functions of two or more units described above can be embodied in one unit. Conversely, the features and functions of one unit described above can be further divided into units embodied by multiple units.

[0161] Furthermore, although the operations of the method of the present application are described in a particular, sequential order, this is not meant to be a limitation, but an additional embodiment, as some of the operations can be performed in a different order, or performed concurrently, that will be apparent to those skilled in the art with the benefit of this disclosure. In addition, some of the operations described herein can be optional, and / or combined with other operations.

[0162] While the preferred embodiments of the application have been described above, it should be understood that they have been presented by way of example only, and not limitation. Numerous changes to the method can be made by those skilled in the art without departing from the true spirit and scope of the application. Thus, the scope of the application should not be limited by the preferred embodiments described herein, but should be given the broadest interpretation of the appended claims and their equivalents.

[0163] Obviously, numerous modifications and variations of the present application are possible in light of the above teachings. It is therefore to be understood that within the scope of the appended claims and their equivalents, the application can be practiced otherwise than as specifically described herein. Industrial Applicability

[0164] By the scheme, the prediction of the negative volume is realized based on the electrocoagulation and electrocision data corresponding to the electrocoagulation and electrocision operation of the doctor in the operation process and the real-time negative volume, the comprehensive utilization of multiple data sources in the hysteroscopy operation early warning process is realized, the prediction of the negative volume is realized comprehensively and accurately, a good data basis for the hysteroscopy operation early warning is provided, and the negative volume early warning for the current operation type can be realized, the early warning effect of the hysteroscopy operation early warning is improved, meanwhile, the hysteroscopy operation early warning method has the prediction ability for the future negative volume change based on the early warning of the predicted negative volume, and the practicability is improved.

Claims

1. A hysteroscopic surgery early warning method, characterized by, The method comprises the following steps: obtaining real-time electrocoagulation and electrocision data, real-time negative balance of uterine distension liquid, and current operation type; wherein the real-time electrocoagulation and electrocision data are real-time electrocoagulation pedal pressure of a hysteroscope electrocoagulation control pedal and real-time electrocision pedal pressure of a hysteroscope electrocision control pedal; the real-time negative balance of uterine distension liquid is determined by real-time inflow weight and real-time outflow weight of the uterine distension liquid; inputting the real-time electrocoagulation and electrocision data and the real-time negative balance into a negative balance prediction model to obtain a predicted negative balance output by the negative balance prediction model; obtaining a first type of warning level corresponding to the current operation type and a first type of warning threshold range corresponding to the first type of warning level based on a mapping relationship between the preset operation type, the first type of warning level, and the first type of warning threshold range; taking the first type of warning level corresponding to the threshold range to which the predicted negative balance belongs as a predicted warning level, and performing a warning operation corresponding to the level of the predicted warning level.

2. The hysteroscopic surgery warning method of claim 1, wherein, After obtaining the real-time electrocoagulation and electrocision data, the real-time negative balance of uterine distension liquid, and the current operation type, the method further comprises the following steps: obtaining historical negative balance of uterine distension liquid; wherein the historical negative balance of uterine distension liquid is determined by historical inflow weight and historical outflow weight of the uterine distension liquid; determining a real-time change value of the negative balance based on the historical negative balance and the real-time negative balance; determining whether the real-time change value is greater than a preset change threshold; if yes, issuing a second type of operation risk prompt in the case that the real-time pedal pressure is not zero.

3. The hysteroscopic surgery warning method of claim 1, wherein, After obtaining the real-time electrocoagulation and electrocision data, the real-time negative balance of uterine distension liquid, and the current operation type, the method further comprises the following steps: obtaining case data of the patient and a type of uterine distension liquid; wherein the case data comprises patient age, current condition of the patient, and basic condition of the patient; inputting the patient age, the current condition of the patient, and the type of uterine distension liquid into a first threshold model evaluation model to obtain a second type of warning level and a threshold range corresponding to the second type of warning level; determining a target threshold range corresponding to the second type of warning level based on the basic condition of the patient and the threshold range corresponding to the second type of warning level; determining a second type of warning level corresponding to the real-time negative balance based on the second type of warning level and the target threshold range corresponding to the second type of warning level, and taking the second type of warning level corresponding to the real-time negative balance as a second type of target warning level; performing a warning operation corresponding to the second type of target warning level based on a corresponding relationship between the second type of warning level and the warning operation. The method further comprises the following steps:

4. The hysteroscopic surgery warning method of claim 3, wherein, determining whether the basic condition of the patient belongs to a preset type of condition; if yes, down-regulating the threshold range corresponding to each second type of warning level, and taking the down-regulated threshold range as the target threshold range corresponding to each second type of warning level; if no, taking the threshold range corresponding to each second type of warning level output by the first threshold model evaluation model as the target threshold range corresponding to each second type of warning level. The method further comprises the following steps:

5. The hysteroscopic surgery warning method of claim 1, wherein, obtaining current intrauterine pressure and mean arterial pressure of the patient; ​ A difference between the intrauterine pressure and the mean arterial pressure is determined as a pressure difference, and a ratio of the pressure difference to the mean arterial pressure is determined as a pressure increase ratio; Based on a preset third type of early warning level and a threshold range corresponding to the third type of early warning level, a third type of early warning level corresponding to the pressure increase ratio is determined, and the third type of early warning level corresponding to the pressure increase ratio is taken as a third type of target early warning level; Based on a corresponding relationship between the third type of early warning level and an early warning operation, an early warning operation corresponding to the third type of target early warning level is performed.

6. The hysteroscopic surgery warning method according to any one of claims 1 to 5, wherein, Further comprising: After the operation is completed, all the electrocoagulation and electrocision data and the negative volume data during the operation are saved; Based on all the electrocoagulation and electrocision data, behavior data during the operation is determined, wherein the behavior data includes a single time length, a total number of times, and a total time length of electrocoagulation and electrocision behavior; An association relationship between the current operation type, the negative volume data, and the behavior data is established, and the association relationship and the current operation type, the uterine distension liquid data, and the behavior data are saved.

7. The hysteroscopic surgery warning method according to claim 6, wherein, Before the real-time electrocoagulation and electrocision data, the real-time negative volume of the uterine distension liquid, and the current operation type are obtained, further comprising: A training set is obtained; wherein the training set includes multiple groups of monitoring data, and each group of monitoring data includes electrocoagulation and electrocision data and negative volume of uterine cavity liquid at the same time; Each group of monitoring data in the training set is input into the negative volume prediction model respectively, and a negative volume prediction result corresponding to each group of monitoring data is obtained; wherein the negative volume prediction result includes a predicted negative volume at a first time after the collection time of the monitoring data; Based on the negative volume prediction result corresponding to each group of monitoring data and a negative volume annotation result, a loss function value is determined; wherein the negative volume annotation result is a true negative volume at the first time after the collection time of the monitoring data; Based on the loss function value, the model parameters of the negative volume prediction model are updated. Comprising:

8. A hysteroscopic surgery early warning device, characterized by, The data acquisition unit is configured to obtain real-time electrocoagulation and electrocision data, real-time negative volume of uterine distension liquid, and current operation type; wherein the real-time electrocoagulation and electrocision data are real-time electrocoagulation pedal pressure of a hysteroscope electrocoagulation control pedal and real-time electrocision pedal pressure of a hysteroscope electrocision control pedal; the real-time negative volume of uterine distension liquid is determined by real-time inflow weight and real-time outflow weight of uterine distension liquid; The data prediction unit is configured to input the real-time electrocoagulation and electrocision data and the real-time negative volume into a negative volume prediction model, and obtain a predicted negative volume output by the negative volume prediction model; The first threshold determination unit is configured to obtain a first type of early warning level corresponding to the current operation type and a first type of early warning threshold range corresponding to the first type of early warning level based on a mapping relationship between preset operation types, first type of early warning levels, and first type of early warning threshold ranges; The first risk prompt unit is configured to take a first type of early warning level corresponding to a threshold range to which the predicted negative volume belongs as a predicted early warning level, and issue a first type of operation risk prompt when the predicted early warning level is higher than a preset level. Further comprising:

9. The hysteroscopic surgery warning device according to claim 8, wherein, ​ The second risk prompting unit is configured to acquire a historical negative amount of the uterine distension liquid; wherein the historical negative amount of the uterine distension liquid is determined by a historical filling weight and a historical outflow weight of the uterine distension liquid; a real-time change value of the negative amount is determined based on the historical negative amount and a real-time negative amount; it is judged whether the real-time change value is greater than a preset change threshold value; if yes, a second type of operation risk prompt is issued in a case that a real-time foot pedal pressure is not zero.

10. The hysteroscopic surgery warning device according to claim 8, wherein, Further comprising: The first early warning unit is configured to acquire case data of a patient and a type of uterine distension liquid; wherein the case data comprises a patient age, a current condition of the patient and a basic condition of the patient; the patient age, the current condition of the patient and the type of uterine distension liquid are input into a first threshold model to obtain a second type of early warning level and a threshold range corresponding to the second type of early warning level; a target threshold range corresponding to the second type of early warning level is determined based on the basic condition of the patient and the threshold range corresponding to the second type of early warning level; a second type of early warning level corresponding to the real-time negative amount is determined based on the second type of early warning level and the target threshold range corresponding to the second type of early warning level, and the second type of early warning level corresponding to the real-time negative amount is taken as a second type of target early warning level; a warning operation corresponding to the second type of target early warning level is executed based on a corresponding relationship between the second type of early warning level and the warning operation. The first pre-unit is specifically configured to:

11. The hysteroscopic surgical warning device according to claim 10, wherein, determine whether the basic condition of the patient belongs to a preset type of condition; if yes, the threshold range corresponding to each second type of early warning level is adjusted downward, and the adjusted threshold range is taken as a target threshold range corresponding to each second type of early warning level; if not, the threshold range corresponding to each second type of early warning level output by the first threshold model is taken as a target threshold range corresponding to each second type of early warning level. Further comprising:

12. The hysteroscopic surgical warning device according to claim 8, wherein, The second early warning unit is configured to acquire a current intrauterine pressure and a mean arterial pressure of a patient; a difference between the intrauterine pressure and the mean arterial pressure is determined as a pressure difference, and a ratio of the pressure difference to the mean arterial pressure is determined as a pressure increase ratio; a third type of early warning level corresponding to the pressure increase ratio is determined based on a preset third type of early warning level and a threshold range corresponding to the third type of early warning level, and the third type of early warning level corresponding to the pressure increase ratio is taken as a third type of target early warning level; a warning operation corresponding to the third type of target early warning level is executed based on a corresponding relationship between the third type of early warning level and the warning operation. Further comprising:

13. The hysteroscopic surgical warning device of any one of claims 8-12, wherein, The data storage processing unit is configured to save all the electrocoagulation and electrocision data and the negative amount data in a surgical process after the surgery is completed; behavior data in the surgical process is determined based on all the electrocoagulation and electrocision data, wherein the behavior data comprises a single time length, a total number and a total time length of electrocision and electrocoagulation behaviors; an association relationship among the current type of surgery, the negative amount data and the behavior data is established, and the association relationship and the current type of surgery, the uterine distension liquid data and the behavior data are saved. Further comprising:

14. The hysteroscopic surgical warning device according to claim 13, wherein, ​ The model training unit is configured to obtain a training set; wherein the training set includes a plurality of groups of monitoring data, wherein each group of monitoring data includes electrocoagulation and electrocision data and negative underfill of uterine fluid at the same time; each group of monitoring data in the training set is input into the negative underfill prediction model to obtain a negative underfill prediction result corresponding to each group of monitoring data; wherein the negative underfill prediction result includes a predicted negative underfill at a first time after the collection time of the monitoring data; a loss function value is determined based on the negative underfill prediction result corresponding to each group of monitoring data and a negative underfill annotation result; wherein the negative underfill annotation result is the actual negative underfill at the first time after the collection time of the monitoring data; and the model parameters of the negative underfill prediction model are updated based on the loss function value.

15. A hysteroscopic surgery early warning system, characterized by, Comprise: a first weight sensor, a second weight sensor, a first pressure sensor, a second pressure sensor, and a processor; the first weight sensor is arranged at the bottom of the uterine distention fluid supply device, and the first weight sensor is configured to collect the inflow weight of the uterine distention fluid; the second weight sensor is arranged at the bottom of the uterine distention fluid outflow storage device, and the second weight sensor is configured to collect the outflow weight of the uterine distention fluid; the first pressure sensor is arranged at the bottom of the hysteroscope electrocoagulation control pedal, and the first pressure sensor is configured to collect the electrocoagulation pedal pressure; the second pressure sensor is arranged at the bottom of the hysteroscope electrocision control pedal, and the second pressure sensor is configured to collect the electrocision pedal pressure; the processor is connected with the first weight sensor, the second weight sensor, the first pressure sensor, and the second pressure sensor respectively; and the processor is configured to execute the computer program corresponding to the hysteroscopy operation early warning method according to any one of claims 1-7.

16. The hysteroscopic surgical warning system of claim 15, wherein, Further comprise: a first flow sensor arranged on the inflow pipeline of the uterine distention fluid and a second flow sensor arranged on the outflow pipeline of the uterine distention fluid; the first flow sensor is configured to collect the inflow flow of the uterine distention fluid; and the second flow sensor is configured to collect the outflow flow of the uterine distention fluid.

17. A computer-readable storage medium, characterized in that, The computer readable storage medium stores computer instructions, and the computer instructions are executed by the processor to implement the hysteroscopy operation early warning method according to any one of claims 1-7.

Images