A system and method for early warning and evaluation of abdominal infection after gastric cancer surgery

Abstract

The present invention discloses a system and method for early warning evaluation of abdominal infection after gastric cancer surgery, which belongs to the technical field of gastric cancer surgery infection, including an information collection module for obtaining the clinical manifestations, pathogenesis characteristics and infection factors of historical infection cases, and also includes obtaining the time of completion of the patient's operation and the time of confirming the infection of the patient, and calculating the time difference from the completion of the operation to the confirmation of the infection, and obtaining the weight of the time difference from the time difference of the historical infection case, and a data management module for grouping the different clinical manifestations and pathogenesis characteristics of the historical infection case. The present invention is based on the observation of the condition, with nurses as the leading factor, to construct an early warning evaluation tool for abdominal infection with high specificity and strong sensitivity, which can identify the precursors of infection early, provide a reliable basis for medical diagnosis, make decision interventions early, shorten the healing time and hospitalization time of patients with anastomotic fistula, reduce the unplanned transfer rate and mortality rate to the ICU, and improve the clinical outcome of patients.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence-assisted medical technology, and in particular to a system and method for early warning and evaluation of abdominal infection after gastric cancer surgery. Background Art

[0002] Gastric cancer is one of the most common malignant tumors. Compared with other abdominal surgeries, postoperative complications of gastric cancer are more unique and complex. The incidence of intra-abdominal infection is 3% to 15%. The main cause of intra-abdominal infection is anastomotic fistula, which is prone to local tissue necrosis, vascular corrosion, bleeding, etc. The mortality rate is as high as 28.5% to 71%.

[0003] Common risk factors for intra-abdominal infection include advanced age, diabetes, malnutrition, etc. Even after improving these factors clinically, intra-abdominal infection may still occur, that is, inevitable intra-abdominal infection, with an incidence of about 4%.

[0004] At present, the judgment of the severity of the patient's condition is mainly influenced by clinical experience and intuition. It is not only difficult to accurately reflect the severity and changes of the disease, but also lacks specificity. In most cases, the "warning signals" of changes in the patient's condition after surgery are difficult to identify. There is a lack of early warning scoring system and quality standards for abdominal infection after gastric cancer surgery, resulting in problems such as poor judgment based on the manual experience of medical staff. It is not easy to detect "potentially critical" patients, but usually before this, most patients have already developed related warning symptoms such as fever and abdominal pain, which delays the best time for treatment and rescue. Summary of the invention

[0005] In view of the above-mentioned problems of low accuracy in the existing assessment of abdominal infection after gastric cancer surgery, the present invention is proposed.

[0006] Therefore, one of the objects of the present invention is to provide a system and method for early warning and evaluation of abdominal infection after gastric cancer surgery, which, based on the observation of the patient's condition, constructs an early warning and evaluation model of abdominal infection with high specificity and sensitivity, and can quickly identify the precursors of abdominal infection, providing a reliable basis for medical staff to make decisions and intervene early. Moreover, based on the condition assessment, the system automatically generates a nursing plan, thereby improving the clinical auxiliary treatment effect of the patient.

[0007] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0008] In one aspect, the present invention provides a system for early warning and evaluation of abdominal infection after gastric cancer surgery, comprising:

[0009] An information collection module is used to obtain relevant data of patients with abdominal infection after surgery for gastric cancer, wherein the relevant data include clinical manifestations, onset characteristics, infection factors, completion time of surgery and time of abdominal infection of patients in historical infection cases, as well as age, gender, preoperative nutrition, presence or absence of complications, duration of surgery, and surgical method of patients, after confirming the time difference of abdominal infection after surgery by the completion time of surgery and time of abdominal infection of patients, weight data matched by the time difference is obtained, and historical associated data of patients with abdominal infection after surgery is obtained according to the time difference of abdominal infection after surgery and the matched weight data, and characteristic data of abdominal infection after gastric cancer surgery is formed after data preprocessing by combining the relevant data and historical associated data of patients with abdominal infection after surgery for gastric cancer;

[0010] A data management module, which is used to respond to the characteristic data of abdominal infection after gastric cancer surgery, cluster different clinical manifestations and onset characteristics of infection cases in historical related data based on a cluster analysis method, obtain characteristic data of different groups, and build an abdominal infection prediction model based on the characteristic data of different groups;

[0011] A data training module, which is used to respond to the characteristic data of different groups of the data management module, train the abdominal infection prediction model through a random forest model, output the training results to determine the final classification results, formulate a graded management plan corresponding to different risk levels according to the classification results, and conduct supervised machine learning training;

[0012] The risk fusion assessment module is used to analyze the clinical manifestations, disease characteristics and infection risks of the current patients based on the relevant data of the current patients with abdominal infection after gastric cancer surgery, obtain the characteristic data of historical infection cases that are susceptible to infection in the historical associated data, and divide the acquired characteristic data of historical infection cases that are susceptible to infection into levels, and at the same time, analyze the change trend data of clinical manifestations at each level and the diverse disease characteristics formed, that is, generate the correlation data between the change trend data at each level and the diverse disease characteristics, respond to the correlation data between the change trend data at each level and the diverse disease characteristics, evaluate the potential risk data of clinical changes at each level, and construct a risk rating table for abdominal infection after gastric cancer surgery based on the potential risk data;

[0013] The control warning module obtains the changing patterns based on the data of different groups, and presets the groups with similar changing patterns as the control group, and the groups with opposite changing patterns as the observation group. When the patient's clinical changes fall within the range of the control group, the system will issue a warning, otherwise, no warning will be issued.

[0014] As a preferred solution of the present invention, the risk fusion assessment module analyzes the current clinical manifestations, disease characteristics and infection risks of the patient, wherein the disease characteristic data include the patient's body temperature, heart rate, abdominal pain and distension, and postoperative white blood cell count, and after the disease characteristic data is imported into the abdominal infection prediction model, the infection risk value is calculated and the risk level is graded to draw the ROC abdominal infection curve. The specific infection risk value is calculated as follows:

[0015] The proportion of different risk levels is obtained based on the age and gender of different patients, and the weight values ​​of the patient's highest body temperature, fastest heart rate, abdominal pain and distension, and postoperative white blood cell count in the characteristic data are calculated under different proportions, and calculated according to the following formula:

[0016] Wherein, k represents the average body temperature °C, average heart rate and average postoperative white blood cell count of the patient measured before and after surgery;

[0017] In the formula, x i represents the patient's highest body temperature, fastest heart rate, abdominal pain and distension, and postoperative white blood cell count data measured for the patient for the i-th time during the first course of medication after surgery. j represents the highest body temperature, fastest heart rate, abdominal pain and distension, and postoperative white blood cell count measured for the patient in the second course of medication after surgery, σ 2 represents the number of complications collected from patients during postoperative medication treatment, exp represents the difference in abdominal pain, abdominal distension and white blood cell count in patients of different age groups during postoperative medication treatment, ∫ 2+1 It shows the changes in the average body temperature ℃, average heart rate and average white blood cell count measured 2+1 times after surgery under different surgical methods.

[0018] As a preferred embodiment of the present invention, the intra-abdominal infection prediction model determines the final classification result by majority voting of the random forest model, formulates graded management plans corresponding to different risk levels according to the classification results and performs machine learning training. Specifically, the machine learning training of the graded management plans corresponding to different risk levels randomly divides the labeled data sets of different risk levels into training sets, validation sets and test sets, and trains each underlying learner with the training set separately, and iterates until all subsets have served as training sets and the training is terminated. At the same time, in the training stage, each underlying learner is trained with the training set separately, and iterates.

[0019] As a preferred embodiment of the present invention, the information collection module further comprises obtaining historical associated data of patients with postoperative abdominal infection according to the time difference of postoperative abdominal infection, combining the relevant data of the patients with abdominal infection and the historical associated data of the patients with postoperative abdominal infection to form characteristic data of postoperative abdominal infection of gastric cancer, wherein the characteristic data of postoperative abdominal infection of gastric cancer comprises the change characteristics of surgical methods, specifically, the surgical methods comprise distal gastrectomy, proximal gastrectomy and total gastrectomy, and collecting the change characteristics of different surgical methods, wherein the change characteristics of the surgical methods comprise the time of appearance of the first bowel sound, the time of first flatus and the time of recovery of intestinal peristalsis after surgery of the patient, and according to the obtained change characteristics, analyzing the associated changes of the time of appearance of bowel sound, the time of first flatus and the time of recovery of intestinal peristalsis with the recovery of body temperature to normal, the recovery of heart rate to normal, the recovery of white blood cell count to normal and the situation of abdominal pain and bloating under different surgical methods, drawing ROC associated change curve at the same time, and calculating the average changes of the highest body temperature ℃, the fastest heart rate, the situation of abdominal pain and bloating and the number of white blood cells in different medication courses under different surgical methods, and calculating the result according to the following formula:

[0020] Among them, ξ i ≥0;

[0021] Wherein, ||w|| represents the white blood cell count measured under different surgical methods, c represents the difference between the white blood cell counts measured under different surgical methods, n represents individual differences, which include age group and gender, and also represents the difference change value of the white blood cell count under different surgical methods under different age groups and genders under the individual differences, ξ i Represents the last white blood cell count data measured in each medication course for the i-th time.

[0022] As a preferred scheme of the present invention, wherein: the information collection module obtains the clinical manifestations, onset characteristics and infection factors of historical infection cases of the same age group, the same gender and different age groups and different genders, and presets the first reference data and the second reference data respectively, the first reference data includes the common clinical manifestations, onset characteristics and infection factors of the same age group and the same gender, the second reference data includes different clinical manifestations, onset characteristics and infection factors of different age groups and different genders, and the first reference data and the second reference data are summarized into a data set.

[0023] As a preferred embodiment of the present invention, the risk fusion assessment module obtains the medication course formulated under different surgical methods, and obtains the patient's historical infection period under the medication course, and obtains the same infection period based on the historical infection period, and presets a key monitoring period based on the same infection period, and presets a monitoring threshold during the key monitoring period. The monitoring threshold includes monitoring of the patient's highest body temperature ℃, fastest heart rate, abdominal pain and distension, postoperative white blood cell count and complications. When the patient's monitoring data exceeds the monitoring threshold, the system issues an infection warning and determines that the patient is infected. Otherwise, the system determines that it is normal.

[0024] As a preferred solution of the present invention, the patient's monitoring data is divided into low, medium and high proximity monitoring data in a step-by-step manner based on the monitoring threshold. When the monitoring data is at low proximity, the system determines that the patient is normal. When the monitoring data is at medium proximity, the system issues an early warning, and the medical team observes changes in the patient's monitoring data. When the monitoring data is at high proximity, the system determines that the patient is infected.

[0025] As a preferred solution of the present invention, a monitoring model is constructed based on the changing pattern of medium-proximity monitoring data to high-proximity monitoring data, and the corresponding monitoring data of different surgical methods, different age groups and different genders are respectively imported into the monitoring model; when the patient's surgical method, age group and gender conform to the changing pattern of the monitoring data in the monitoring model, the system determines that the monitoring data will develop towards high-proximity monitoring data, otherwise, no determination is made.

[0026] In another aspect, the present invention provides a method for an early warning evaluation system for abdominal infection after gastric cancer surgery, comprising the following steps:

[0027] Step S10: collecting clinical manifestations, onset characteristics and infection factors of historical infection cases, and collecting the time when the patient's surgery was completed and the time when the patient's infection was confirmed, and calculating the time difference between the completion of the surgery and the confirmation of the infection, and calculating the weight of the time difference compared with the time difference of historical infection cases;

[0028] Step S20: grouping different clinical manifestations and onset characteristics of historical infection cases, obtaining characteristic data of different groups, and constructing an abdominal infection prediction model based on the characteristic data of different groups, wherein the characteristic data of different groups include the patient's highest body temperature, fastest heart rate, abdominal pain and distension, and postoperative white blood cell count, and importing the characteristic data into the abdominal infection prediction model;

[0029] Step S30: presetting an infection threshold based on the characteristic data of different groups, and performing infection risk analysis when the clinical manifestation data of the postoperative patient reaches the infection threshold;

[0030] Step S40: obtaining relevant data information of patients in different groups, including the age, gender, preoperative nutrition, presence or absence of complications, and operation time and method of the patients, and analyzing the clinical manifestations and disease characteristics of the patients under different data information conditions, as well as the infection risks caused;

[0031] Step S50: based on the data of different groups, the changing rules are obtained, and the groups with similar changing rules are preset as the control group, and the groups with opposite changing rules are preset as the observation group. When the clinical changes of the patient belong to the range of the control group, the system issues an early warning, otherwise, no early warning is issued;

[0032] Step S60: Obtaining data features of susceptible individuals in historical infection cases, and dividing the obtained relevant data information into hierarchical levels;

[0033] Step S70: Analyze the changing patterns of clinical manifestations at each level and the diverse disease characteristics formed, and obtain the correlation between each changing pattern and the diverse disease characteristics. At the same time, obtain the evolution pattern of clinical changing patterns at each level to susceptible data, analyze the potential risks of clinical changes at each level, and construct a risk rating table based on the potential risks.

[0034] Beneficial effects of the present invention: Based on the observation of the patient's condition, the present invention constructs an early warning and evaluation model for abdominal infection with high specificity and sensitivity, wherein the abdominal infection prediction model adopts a machine learning algorithm, and the model performs feature selection on the input data to screen out features that have an important impact on the abdominal infection after gastric cancer surgery, which helps to improve the prediction ability and accuracy of the model, and can quickly identify the precursors of abdominal infection, providing a reliable basis for medical staff to make decisions and intervene early. In addition, based on the condition assessment, the system automatically generates a nursing plan, which improves the clinical auxiliary treatment effect of the patient. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative labor. Among them:

[0036] Figure 1 This is a schematic diagram of the modular structure of the early warning and evaluation system for abdominal infection after gastric cancer surgery in an embodiment of the present invention;

[0037] Figure 2 A schematic diagram of a system flow of an embodiment of the present invention;

[0038] Figure 3 is a correlation change curve diagram of an embodiment of the present invention;

[0039] Figure 4 A schematic diagram of the hardware structure used in the system program of the embodiment of the present invention;

[0040] Numbers in the figure: 110 - information collection module; 120 - data management module; 130 - data analysis module; 140 - risk fusion assessment module; 150 - comparison and warning module. DETAILED DESCRIPTION

[0041] In order to make the purpose, technical solution and advantages of the embodiment of the present invention clearer, the technical solution of the embodiment of the present invention will be clearly and completely described below in conjunction with the drawings of the embodiment of the present invention. Obviously, the described embodiment is a part of the embodiment of the present invention, not all of the embodiments. Based on the described embodiment of the present invention, all other embodiments obtained by ordinary technicians in this field belong to the scope of protection of the present invention.

[0042] Reference Figures 1 to 4 , is an embodiment of the present invention, which provides a system for early warning and evaluation of abdominal infection after gastric cancer surgery, comprising:

[0043] The information collection module 110 is used to obtain relevant data of patients with abdominal infection after surgery for gastric cancer, wherein the relevant data include clinical manifestations, onset characteristics, infection factors, completion time of surgery and time of abdominal infection of patients in history of infection cases, as well as age, gender, preoperative nutrition, presence or absence of complications, duration of surgery, and surgical method of patients, after confirming the time difference of abdominal infection after surgery by the completion time of surgery and time of abdominal infection of patients, weight data matched by the time difference is obtained, and historical associated data of patients with abdominal infection after surgery is obtained according to the time difference of abdominal infection after surgery and the matched weight data, and characteristic data of abdominal infection after gastric cancer surgery is formed after data preprocessing by combining the relevant data and historical associated data of patients with abdominal infection after surgery for gastric cancer.

[0044] It should be noted that in this embodiment, under the premise of ensuring the legal data acquisition authority and complying with medical data protection regulations, the present invention can obtain medical records and clinical data through the electronic health record system (EHR), obtain real-time data through medical equipment, or obtain laboratory result data from the laboratory database, etc. The present invention does not make too many restrictions on the specific collection method, as long as the corresponding data can be obtained; in addition, when obtaining the clinical manifestations, onset characteristics and infection factors of historical infection cases, it also includes obtaining the clinical manifestations, onset characteristics and infection factors of historical infection cases of the same age group, the same gender and different age groups and different genders;

[0045] It should be emphasized in this embodiment that the information collection module 110 can preset the first reference data and the second reference data respectively, the first reference data includes the common clinical manifestations, disease characteristics and infection factors of the same age group and the same gender, and the second reference data includes different clinical manifestations, disease characteristics and infection factors of different age groups and different genders, and the first reference data and the second reference data are summarized into a data set. When the postoperative clinical manifestations of the patient in the future time period meet the range of the first reference data, the system issues an infection warning and determines that the patient is infected. When the postoperative clinical manifestations of the patient in the future time period meet the range of the second reference data, the system also issues an infection warning, but does not determine that the patient is infected;

[0046] The data management module 120 is used to respond to the characteristic data of the abdominal infection after gastric cancer surgery, cluster different clinical manifestations and onset characteristics in the infection cases of historical correlation data based on the cluster analysis method, obtain characteristic data of different groups, and build an abdominal infection prediction model based on the characteristic data of different groups. In this embodiment, different clinical manifestations and onset characteristics of historical infection cases are grouped, and characteristic data of different groups are obtained. At the same time, infection thresholds are preset based on the characteristic data of different groups. When the clinical manifestation data of the postoperative patient reaches the infection threshold, the system determines that the patient is infected. Otherwise, the system determines that the patient is not infected, and builds an IAI-EWS early warning model based on the characteristic data of different groups. The IAI-EWS early warning model here is an abdominal infection prediction model;

[0047] Specifically, in this embodiment, the characteristic data of different groups include the patient's highest body temperature, fastest heart rate, abdominal pain and bloating, and postoperative white blood cell count, and the characteristic data are imported into the IAI-EWS early warning model, and the risk level is graded at the same time, where 0 to 4 points are low risk, 5 to 6 points are medium risk, 7 to 10 points are high risk, and 11 to 12 points are extremely high risk.

[0048] A data training module 130, which is used to respond to the characteristic data of different groups of the data management module, train the abdominal infection prediction model through a random forest model, output the training results to determine the final classification results, formulate a graded management plan corresponding to different risk levels according to the classification results, and perform supervised machine learning training;

[0049] The risk fusion assessment module 140 is used to analyze the clinical manifestations, disease characteristics and infection risks of the current patients based on the relevant data of the current patients with abdominal infection after gastric cancer surgery, obtain the characteristic data of the historical infection cases that are susceptible to infection in the historical associated data, and divide the acquired characteristic data of the historical infection cases that are susceptible to infection into levels, and at the same time, analyze the change trend data of the clinical manifestations at each level and the formed diverse disease characteristics, that is, generate the correlation data between the change trend data at each level and the diverse disease characteristics, respond to the correlation data between the change trend data at each level and the diverse disease characteristics, evaluate the potential risk data of the clinical changes at each level, and construct a risk rating table for abdominal infection after gastric cancer surgery based on the potential risk data;

[0050] This embodiment analyzes the current clinical manifestations, disease characteristics and infection risks of patients, wherein the disease characteristic data include the patient's body temperature, heart rate, abdominal pain and distension, and postoperative white blood cell count, and obtains the proportion of different risk levels based on the age and gender of different patients, and calculates the weight values ​​of the patient's highest body temperature ℃, fastest heart rate, abdominal pain and distension, and postoperative white blood cell count in the characteristic data under different proportions, and imports the disease characteristic data into the abdominal infection prediction model, calculates the infection risk value, and draws the ROC abdominal infection curve after the risk level is graded. The specific infection risk value calculation is as follows:

[0051] Wherein, k represents the average body temperature °C, average heart rate and average postoperative white blood cell count of the patient measured before and after surgery;

[0052] In the formula, x i represents the patient's highest body temperature, fastest heart rate, abdominal pain and distension, and postoperative white blood cell count data measured for the patient for the i-th time during the first course of medication after surgery. j represents the highest body temperature, fastest heart rate, abdominal pain and distension, and postoperative white blood cell count measured for the patient in the second course of medication after surgery, σ 2 represents the number of complications collected from patients during postoperative medication treatment, exp represents the difference in abdominal pain, abdominal distension and white blood cell count in patients of different age groups during postoperative medication treatment, ∫ 2+1 It shows the changes in the average body temperature ℃, average heart rate and average white blood cell count measured 2+1 times after surgery under different surgical methods.

[0053] This embodiment further illustrates that the information collection module 110 also includes obtaining historical associated data of patients with postoperative abdominal infection according to the time difference of postoperative abdominal infection, combining the relevant data of the patients with abdominal infection and the historical associated data of the patients with postoperative abdominal infection to form characteristic data of postoperative abdominal infection of gastric cancer, wherein the characteristic data of postoperative abdominal infection of gastric cancer includes the change characteristics of surgical methods, specifically, the surgical methods include distal gastrectomy, proximal gastrectomy and total gastrectomy, and collecting the change characteristics of different surgical methods, wherein the change characteristics of the surgical methods include the time of appearance of the first bowel sound, the time of first flatus and the time of recovery of intestinal peristalsis after surgery of the patient, and according to the obtained change characteristics, analyzing the correlation changes of the time of appearance of bowel sound, the time of first flatus and the time of recovery of intestinal peristalsis with the recovery of body temperature, the recovery of heart rate, the recovery of white blood cell count and abdominal pain and bloating under different surgical methods, and drawing the ROC correlation change curve, and calculating the average changes of the patient's highest body temperature ℃, the fastest heart rate, the abdominal pain and bloating and the number of white blood cells in different medication courses under different surgical methods, according to the following formula:

[0054] Among them, ξ i ≥0;

[0055] Wherein, ||w|| represents the white blood cell count measured under different surgical methods, c represents the difference between the white blood cell counts measured under different surgical methods, n represents individual differences, which include age group and gender, and also represents the difference change value of the white blood cell count under different surgical methods under different age groups and genders under the individual differences, ξ i Represents the last white blood cell count data measured in each medication course for the i-th time.

[0056] Among them, the ROC abdominal infection curve is used as the performance evaluation index of the meta-learner, the user's wound grade data and the model performance evaluation index are input into the bottom-level learner, the output result of each bottom-level learner is input into the top-level meta-learner, and the output result is obtained through the meta-learner.

[0057] In this embodiment, preferably, a hierarchical management plan corresponding to different hazard levels can be formulated according to the IAI-EWS early warning model through evidence-based and / or Delphi method, where:

[0058] Low risk level: Conduct and update the IAI-EWS warning model assessment once a day;

[0059] Moderate risk level: Evaluate twice a day, notify the senior nurse to re-evaluate, report to the doctor and head nurse, and check on the patients every hour;

[0060] High-risk and extremely high-risk patients: Evaluate every 4 hours, report to the doctor and head nurse immediately, and give the patient oral medication or emergency CT scan according to the doctor's instructions. At the same time, check on the patient every hour. If the score continues to increase, the medical team will discuss the treatment plan or transfer to the ICU for treatment;

[0061] The control warning module 150 obtains the changing patterns based on the data of different groups, and presets the groups with similar changing patterns as the control group, and presets the groups with opposite changing patterns as the observation group. When the patient's clinical changes fall within the range of the control group, the system issues a warning, otherwise, no warning is issued.

[0062] On the basis of the above, this embodiment obtains the medication course formulated for different surgical methods, and obtains the patient's historical infection period under the medication course. At the same time, the same infection period is obtained based on the historical infection period, and the key monitoring period is preset based on the same infection period. The monitoring threshold is preset during the key monitoring period. The monitoring threshold includes the monitoring of the patient's highest body temperature ℃, fastest heart rate, abdominal pain and distension, postoperative white blood cell count and complications. When the patient's monitoring data exceeds the monitoring threshold, the system issues an infection warning and determines that the patient is infected. Otherwise, the system determines that it is normal.

[0063] Among them, this embodiment divides the patient's monitoring data into low, medium and high proximity monitoring data in a step-by-step manner based on the monitoring threshold. When the monitoring data is at low proximity, the system determines that the patient is normal. When the monitoring data is at medium proximity, the system issues an early warning, and the medical team observes changes in the patient's monitoring data. When the monitoring data is at high proximity, the system determines that the patient is infected.

[0064] What needs to be explained in this embodiment is that, on the basis of the above, a monitoring model is constructed based on the changing pattern of medium-proximity monitoring data to high-proximity monitoring data, and the corresponding monitoring data of different surgical methods and different age groups and different genders are respectively imported into the monitoring model. When the patient's surgical method, age group and gender conform to the changing pattern of the monitoring data in the monitoring model, the system determines that the monitoring data will develop towards high-proximity monitoring data, otherwise, it will not make a determination.

[0065] Based on the above, this application constructs a highly specific and sensitive abdominal infection early warning assessment tool through observation of the condition, with nurses as the main body. It can identify the signs of infection early, provide a reliable basis for medical diagnosis, and make early decision-making interventions, shorten the fistula healing time and hospitalization time of patients with anastomotic fistulas, reduce the unplanned ICU transfer rate and mortality rate, and improve the clinical outcomes of patients.

[0066] like Figure 3As shown, this embodiment combines the above-mentioned abdominal infection early warning and evaluation system after gastric cancer surgery and also proposes a working method of the system, as follows:

[0067] Step S10: collecting clinical manifestations, onset characteristics and infection factors of historical infection cases, and collecting the time when the patient's surgery was completed and the time when the patient's infection was confirmed, and calculating the time difference between the completion of the surgery and the confirmation of the infection, and calculating the weight of the time difference compared with the time difference of historical infection cases;

[0068] Step S20: Group different clinical manifestations and onset characteristics of historical infection cases, obtain characteristic data of different groups, and build an IAI-EWS early warning model based on the characteristic data of different groups. The characteristic data of different groups include the patient's highest body temperature, fastest heart rate, abdominal pain and distension, and postoperative white blood cell count, and import the characteristic data into the IAI-EWS early warning model;

[0069] Step S30: presetting an infection threshold based on the characteristic data of different groups, and performing infection risk analysis when the clinical manifestation data of the postoperative patient reaches the infection threshold;

[0070] Step S40: obtaining relevant data information of patients in different groups, including the age, gender, preoperative nutrition, presence or absence of complications, and operation time and method of the patients, and analyzing the clinical manifestations and disease characteristics of the patients under different data information conditions, as well as the infection risks caused;

[0071] Step S50: based on the data of different groups, the changing rules are obtained, and the groups with similar changing rules are preset as the control group, and the groups with opposite changing rules are preset as the observation group. When the clinical changes of the patient belong to the range of the control group, the system issues an early warning, otherwise, no early warning is issued;

[0072] Step S60: Obtain the data features of susceptible cases in historical infection cases, and divide the obtained relevant data information into levels as follows:

[0073] Low risk level: Conduct IAI-EWS early warning model assessment once a day and update the IAI-EWS early warning model assessment;

[0074] Moderate risk level: Evaluate twice a day, notify the senior nurse to re-evaluate, report to the doctor and head nurse, and check on the patients every hour;

[0075] High-risk and extremely high-risk patients: Evaluate every 4 hours, report to the doctor and head nurse immediately, and give the patient oral medication or emergency CT scan according to the doctor's instructions. At the same time, check on the patient every hour. If the score continues to increase, the medical team will discuss the treatment plan or transfer to the ICU for treatment;

[0076] Step S70: Analyze the changing patterns of clinical manifestations at each level and the diverse disease characteristics formed, and obtain the correlation between each changing pattern and the diverse disease characteristics. At the same time, obtain the evolution pattern of clinical changing patterns at each level to susceptible data, analyze the potential risks of clinical changes at each level, and construct a risk rating table based on the potential risks.

[0077] Figure 4 The schematic diagram of the hardware structure used in applying the early warning and evaluation system program for abdominal infection after gastric cancer surgery according to the embodiment method of the present application.

[0078] The early warning and evaluation system for abdominal infection after gastric cancer surgery may include a processor 41 and a memory 42 storing computer program instructions. It is understood that the structure illustrated in the embodiment of the present application does not constitute a specific limitation on the electronic device. In other embodiments of the present application, the early warning and evaluation system for abdominal infection after gastric cancer surgery may include more or fewer components than shown in the figure, or combine certain components, or split certain components, or arrange the components differently. The illustrated components may be implemented in hardware, software, or a combination of software and hardware.

[0079] Specifically, the processor 41 may include a central processing unit (CPU), or an application specific integrated circuit (ASIC), or may be configured to implement one or more integrated circuits of the embodiments of the present application.

[0080] The memory 42 may include a large capacity memory for data or instructions. By way of example and not limitation, the memory 42 may include a hard disk drive (HDD), a floppy disk drive, a solid state drive (OPENCV), a flash memory, an optical disk, a magneto-optical disk, a magnetic tape, or a universal serial bus (USB) drive, or a combination of two or more of these.

[0081] The memory 42 can be used to store or cache various data files that need to be processed and / or used for communication, as well as possible computer program instructions executed by the processor 41; the processor 41 implements the analysis and processing method in the above embodiment by reading and executing the computer program instructions stored in the memory 42.

[0082] In some of the embodiments, the electronic device may further include a communication interface 43 and a bus 40. Figure 4 As shown, the processor 41, the memory 42, and the communication interface 43 are connected via a bus 40 and communicate with each other.

[0083] In summary, the present invention constructs an early warning and evaluation model for abdominal infection with high specificity and sensitivity based on the observation of the patient's condition. The abdominal infection prediction model adopts a machine learning algorithm. The model will perform feature selection on the input data and screen out features that have an important impact on the abdominal infection after gastric cancer surgery. This helps to improve the predictive ability and accuracy of the model, and at the same time can quickly identify the precursors of abdominal infection, providing medical staff with a reliable basis for early decision-making intervention. In addition, based on the condition assessment, the system automatically generates a nursing plan, which improves the clinical auxiliary treatment effect of the patient.

[0084] The following is a code schematic of some modules of this embodiment, as follows:

[0085] Information collection module 110:

[0086] import numpy as np

[0087] import matplotlib.pyplot as plt

[0088] from sklearn import datasets

[0089] from sklearn.preprocessing import StandardScaler

[0090] from sklearn.svm import SVC

[0091] #Generate the weight of the time difference between the completion of surgery and the confirmation of infection and the time difference of the historical infection cases

[0092] X,y=datasets.make_classification(n_samples=200,n_features=2,n_informative=2,

[0093] n_redundant=0, n_clusters_per_class=1, weights=[0.1,0.9],

[0094] flip_y=0,random_state=42)

[0095] #Data Standardization

[0096] scaler = StandardScaler()

[0097] X = scaler.fit_transform(X)

[0098] Data management module 120:

[0099] svm=SVC(kernel='linear', class_weight='balanced', C=1.0)svm.fit(X,y)

[0100] #Preset infection threshold and decision boundary

[0101] plt.scatter(X[:,0],X[:,1],c=y,cmap=plt.cm.Paired)

[0102] ax = plt.gca()

[0103] xlim = ax.get_xlim()

[0104] ylim = ax.get_ylim()

[0105] #Build IAI-EWS warning model based on characteristic data of different groups

[0106] xx=np.linspace(xlim[0],xlim[1],30)

[0107] yy=np.linspace(ylim[0],ylim[1],30)

[0108] YY,XX=np.meshgrid(yy,xx)

[0109] xy=np.vstack([XX.ravel(),YY.ravel()]).TZ=svm.decision_function(xy).reshape(XX.shape)

[0110] Data Analysis Module 130:

[0111] import numpy as np

[0112] import matplotlib.pyplot as plt

[0113] from sklearn.svm import SVR

[0114] #Get relevant data information of patients in different groups

[0115] X=np.sort(5*np.random.rand(200,1),axis=0)

[0116] y = np.sin(X).ravel()

[0117] #Add the patient's age, gender, preoperative nutrition, comorbidities, and surgery time and method y[::5]+=3*(0.5-np.random.rand(40))

[0118] #Risk of infection caused

[0119] svr_rbf=SVR(kernel='rbf', C=100, gamma=0.1, epsilon=0.1) svr_rbf.fit(X,y)

[0120] #predict

[0121] X_test=np.linspace(0,5,100)[:,None]

[0122] y_pred=svr_rbf.predict(X_test)

[0123] Risk Fusion Assessment Module 140:

[0124] import numpy as np

[0125] import matplotlib.pyplot as plt

[0126] from sklearn import datasets

[0127] from sklearn.preprocessing import StandardScaler

[0128] from sklearn.svm import SVC

[0129] #Get the data features of susceptible infections in historical infection cases

[0130] X,y=datasets.make_classification(n_samples=200,n_features=2,n_informative=2,

[0131] n_redundant=0, n_clusters_per_class=1, weights=[0.1,0.9],

[0132] flip_y=0,random_state=42)

[0133] #Data Standardization

[0134] scaler = StandardScaler()

[0135] X = scaler.fit_transform(X)

[0136] #Model training

[0137] svm=SVC(kernel='linear', class_weight='balanced', C=1.0)

[0138] svm.fit(X,y)

[0139] #Build a risk rating table based on potential risks

[0140] plt.scatter(X[:,0],X[:,1],c=y,cmap=plt.cm.Paired)

[0141] ax = plt.gca()

[0142] xlim = ax.get_xlim()

[0143] ylim = ax.get_ylim()

[0144] Reference to early warning module 150:

[0145] import numpy as np

[0146] import matplotlib.pyplot as plt

[0147] from sklearn import datasets

[0148] from sklearn.preprocessing import StandardScaler

[0149] from sklearn.svm import SVC

[0150] #Acquire the changing rules based on data from different groups

[0151] X,y=datasets.make_classification(n_samples=100,n_features=2,n_informative=2,

[0152] n_redundant=0,n_clusters_per_class=1,random_state=42)

[0153] #Data Standardization

[0154] scaler = StandardScaler()

[0155] X = scaler.fit_transform(X)

[0156] #Default control group

[0157] svm = SVC (kernel = 'linear')

[0158] #Preset observation group

[0159] for i in range(10):

[0160] X_temp,y_temp=datasets.make_classification(n_samples=1,n_features=2,n_informative=2,

[0161] n_redundant=0,n_clusters_per_class=1,random_state=42+i)

[0162] X_temp=scaler.transform(X_temp)

[0163] svm.fit(X_temp,y_temp)

[0164] Based on the above, this embodiment constructs a highly specific and sensitive abdominal infection early warning assessment model based on the patient's condition observation, which can quickly identify the precursors of abdominal infection, provide a reliable basis for medical staff, and make decision-making interventions as early as possible. Moreover, based on the condition assessment, the system automatically generates a nursing plan, which improves the clinical auxiliary treatment effect of the patient, which is specifically reflected as follows:

[0165] First, through the scientific and rigorous research design of the software or hardware of this embodiment, a nurse-led, condition-observation-based, high-specificity, and high-sensitivity abdominal infection early warning assessment tool can be constructed, which can identify the precursors of infection early, provide a reliable basis for medical diagnosis, and make decision interventions early, shorten the fistula healing time and hospitalization time of patients with anastomotic fistulas, reduce the unplanned ICU transfer rate and mortality rate, and improve the clinical outcomes of patients.

[0166] Secondly, the abdominal infection early warning scoring system can be embedded in the intelligent information system. Based on the condition assessment, the system automatically generates a nursing plan to facilitate nursing staff to implement precise nursing measures. At the same time, it automatically evaluates the nursing effect, achieves early warning, helps nurses detect changes in the patient's condition in a timely manner, and provides support for nursing decisions.

[0167] Third, a nurse-led early warning scoring tool for abdominal infection after gastric cancer surgery can be developed to help clinical nurses detect abdominal infection in patients earlier, more conveniently and quickly, give full play to the subjective initiative of nursing staff, and reflect the professionalism and scientific nature of nursing work. Combined with the SBAR standardized communication model, the quality of communication can be improved to win more treatment time for patients.

[0168] In the above embodiments, it can be implemented in whole or in part by software, hardware, firmware or any combination thereof. When implemented using software, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the process or function according to the present application is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium.

[0169] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present application. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine different embodiments or examples described in this specification and the features of different embodiments or examples, unless they are contradictory.

[0170] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include at least one of the features. In the description of this application, the meaning of "plurality" is two or more, unless otherwise clearly and specifically defined.

[0171] Any process or method description in the flow chart or otherwise described herein can be understood to represent a module, fragment or portion of a code including one or more executable instructions for implementing the steps of a specific logical function or process. And the scope of the preferred embodiment of the present application includes other implementations, in which the functions may not be performed in the order shown or discussed, including in a substantially simultaneous manner or in a reverse order according to the functions involved.

[0172] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, which can be embodied in any computer-readable medium for use by an instruction execution system, apparatus or device (such as a computer-based system, a system including a processor or other system that can fetch instructions from an instruction execution system, apparatus or device and execute instructions), or used in combination with these instruction execution systems, apparatuses or devices.

[0173] It should be understood that the various parts of the present application can be implemented with hardware, software, firmware or a combination thereof. In the above embodiments, multiple steps or methods can be implemented with software or firmware stored in a memory and executed by a suitable instruction execution system. All or part of the steps of the above embodiment method can be completed by instructing the relevant hardware through a program, which can be stored in a computer-readable storage medium, and when the program is executed, it includes one of the steps of the method embodiment or a combination thereof.

[0174] In addition, each functional unit in each embodiment of the present application can be integrated into a processing module, or each unit can exist physically separately, or two or more units can be integrated into one module. The above-mentioned integrated module can be implemented in the form of hardware or in the form of a software functional module. If the above-mentioned integrated module is implemented in the form of a software functional module and sold or used as an independent product, it can also be stored in a computer-readable storage medium. The storage medium can be a read-only memory, a disk or an optical disk, etc.

[0175] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of various changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A system for early warning and evaluation of abdominal infection after gastric cancer surgery, characterized in that: include: An information collection module is used to obtain relevant data of patients with abdominal infection after surgery for gastric cancer, wherein the relevant data include clinical manifestations, onset characteristics, infection factors, completion time of surgery and time of abdominal infection of patients in historical infection cases, as well as age, gender, preoperative nutrition, presence or absence of complications, duration of surgery, and surgical method of patients, after confirming the time difference of abdominal infection after surgery by the completion time of surgery and time of abdominal infection of patients, weight data matched by the time difference is obtained, and historical associated data of patients with abdominal infection after surgery is obtained according to the time difference of abdominal infection after surgery and the matched weight data, and characteristic data of abdominal infection after gastric cancer surgery is formed after data preprocessing by combining the relevant data and historical associated data of patients with abdominal infection after surgery for gastric cancer; A data management module, which is used to respond to the characteristic data of abdominal infection after gastric cancer surgery, cluster different clinical manifestations and onset characteristics of infection cases in historical related data based on a cluster analysis method, obtain characteristic data of different groups, and build an abdominal infection prediction model based on the characteristic data of different groups; A data training module, which is used to respond to the characteristic data of different groups of the data management module, train the abdominal infection prediction model through a random forest model, output the training results to determine the final classification results, formulate a graded management plan corresponding to different risk levels according to the classification results, and conduct supervised machine learning training; The risk fusion assessment module is used to analyze the clinical manifestations, disease characteristics and infection risks of the current patients based on the relevant data of the current patients with abdominal infection after gastric cancer surgery, obtain the characteristic data of historical infection cases that are susceptible to infection in the historical associated data, and divide the acquired characteristic data of historical infection cases that are susceptible to infection into levels, and at the same time, analyze the change trend data of clinical manifestations at each level and the diverse disease characteristics formed, that is, generate the correlation data between the change trend data at each level and the diverse disease characteristics, respond to the correlation data between the change trend data at each level and the diverse disease characteristics, evaluate the potential risk data of clinical changes at each level, and construct a risk rating table for abdominal infection after gastric cancer surgery based on the potential risk data; The control warning module obtains the changing patterns based on the data of different groups, and presets the groups with similar changing patterns as the control group, and the groups with opposite changing patterns as the observation group. When the patient's clinical changes fall within the range of the control group, the system will issue a warning, otherwise, no warning will be issued.

2. The early warning and evaluation system for abdominal infection after gastric cancer surgery according to claim 1, characterized in that: The risk fusion assessment module analyzes the current patient's clinical manifestations, disease characteristics, and infection risks caused, wherein the disease characteristic data include the patient's body temperature, heart rate, abdominal pain and distension, and postoperative white blood cell count, and imports the disease characteristic data into the abdominal infection prediction model, calculates the infection risk value, and draws the ROC abdominal infection curve after performing risk grading. The specific infection risk value calculation is as follows: The proportion of different risk levels is obtained based on the age and gender of different patients, and the weight values ​​of the patient's highest body temperature, fastest heart rate, abdominal pain and distension, and postoperative white blood cell count in the characteristic data are calculated under different proportions, and calculated according to the following formula: Wherein, k represents the average body temperature °C, average heart rate and average postoperative white blood cell count of the patient measured before and after surgery; In the formula, x i represents the patient's highest body temperature, fastest heart rate, abdominal pain and distension, and postoperative white blood cell count data measured for the patient for the i-th time during the first course of medication after surgery. j represents the highest body temperature, fastest heart rate, abdominal pain and distension, and postoperative white blood cell count measured for the patient in the second course of medication after surgery, σ 2 represents the number of complications collected from patients during postoperative medication treatment, exp represents the difference in abdominal pain, abdominal distension and white blood cell count in patients of different age groups during postoperative medication treatment, ∫ 2+1 It shows the changes in the average body temperature ℃, average heart rate and average white blood cell count measured 2+1 times after surgery under different surgical methods.

3. The early warning and evaluation system for abdominal infection after gastric cancer surgery according to claim 2, characterized in that: The intra-abdominal infection prediction model determines the final classification result by majority voting of the random forest model, formulates graded management plans corresponding to different risk levels according to the classification results, and performs machine learning training. Specifically, the machine learning training of the graded management plans corresponding to different risk levels randomly divides the labeled data sets of different risk levels into training sets, validation sets, and test sets, and trains each underlying learner with the training set, and iterates until all subsets have served as training sets and the training is terminated. At the same time, in the training stage, each underlying learner is trained with the training set, and iterates; By building multiple decision trees and averaging or voting their results, the robustness and accuracy of the model can be improved. These different types of underlying learners can complement and enhance each other to improve the performance and accuracy of the model. At the same time, combining different types of underlying learners can also reduce the bias and variance of individual models and improve the stability and robustness of the model.

4. The early warning and evaluation system for abdominal infection after gastric cancer surgery according to claim 1, characterized in that: The information collection module also includes obtaining historical associated data of patients with postoperative abdominal infection according to the time difference of postoperative abdominal infection, combining the relevant data of the patients with abdominal infection and the historical associated data of the patients with postoperative abdominal infection to form characteristic data of postoperative abdominal infection after gastric cancer surgery, wherein the characteristic data of postoperative abdominal infection after gastric cancer surgery includes the change characteristics of surgical methods, specifically, the surgical methods include distal gastrectomy, proximal gastrectomy and total gastrectomy, and collecting the change characteristics of different surgical methods, wherein the change characteristics of the surgical methods include the time of the first appearance of bowel sounds, the time of the first flatus and the time of recovery of intestinal peristalsis after surgery, and analyzing the associated changes of the time of the appearance of bowel sounds, the time of the first flatus and the time of recovery of intestinal peristalsis and abdominal pain and bloating under different surgical methods according to the acquired change characteristics, and drawing the ROC associated change curve, and calculating the average value change of the number of white blood cells of patients in different medication courses under different surgical methods, which is calculated according to the following formula: Among them, ξ i ≥0; Wherein, ||w|| represents the white blood cell count measured under different surgical methods, c represents the difference between the white blood cell counts measured under different surgical methods, n represents individual differences, which include age group and gender, and also represents the difference change value of the white blood cell count under different surgical methods under different age groups and genders under the individual differences, ξ i Represents the last white blood cell count data measured in each medication course for the i-th time.

5. The early warning and evaluation system for abdominal infection after gastric cancer surgery as claimed in claim 1, characterized in that: The information collection module obtains the clinical manifestations, onset characteristics and infection factors of historical infection cases of the same age group and the same gender and different age groups and different genders, and presets first reference data and second reference data respectively. The first reference data includes common clinical manifestations, onset characteristics and infection factors of the same age group and the same gender, and the second reference data includes different clinical manifestations, onset characteristics and infection factors of different age groups and different genders, and the first reference data and the second reference data are summarized into a data set.

6. The early warning and evaluation system for abdominal infection after gastric cancer surgery according to claim 1, characterized in that: The risk fusion assessment module obtains the medication courses formulated under different surgical methods, and obtains the patient's historical infection period under the medication course, and at the same time obtains the same infection period based on the historical infection period, and presets the key monitoring period based on the same infection period, and presets the monitoring threshold during the key monitoring period. The monitoring threshold includes the monitoring of the patient's highest body temperature, fastest heart rate, abdominal pain and distension, postoperative white blood cell count and complications.

7. A system for early warning and evaluation of abdominal infection after gastric cancer surgery as claimed in claim 6, characterized in that: The patient's monitoring data is divided into low, medium and high proximity monitoring data in a step-by-step manner based on the monitoring threshold.

8. The early warning and evaluation system for abdominal infection after gastric cancer surgery according to claim 7, characterized in that: A monitoring model is constructed based on the changing rules of medium-proximity monitoring data to high-proximity monitoring data, and corresponding monitoring data of different surgical methods and different age groups and different genders are respectively imported into the monitoring model.

9. The method of the early warning evaluation system for abdominal infection after gastric cancer surgery as claimed in claim 1, characterized in that: The following steps are involved: step S10: Collect the clinical manifestations, onset characteristics and infection factors of historical infection cases, and collect the time of completion of the patient's surgery and the time of confirmation of the patient's infection. At the same time, calculate the time difference from the completion of the surgery to the confirmation of the infection, and calculate the weight of the time difference compared with the time difference of historical infection cases; Step S20: grouping different clinical manifestations and onset characteristics of historical infection cases, obtaining characteristic data of different groups, and constructing an abdominal infection prediction model based on the characteristic data of different groups, wherein the characteristic data of different groups include the patient's highest body temperature, fastest heart rate, abdominal pain and distension, and postoperative white blood cell count, and importing the characteristic data into the abdominal infection prediction model; Step S30: presetting an infection threshold based on the characteristic data of different groups, and performing infection risk analysis when the clinical manifestation data of the postoperative patient reaches the infection threshold; Step S40: obtaining relevant data information of patients in different groups, including the age, gender, preoperative nutrition, presence or absence of complications, and operation time and method of the patients, and analyzing the clinical manifestations and disease characteristics of the patients under different data information conditions, as well as the infection risks caused; Step S50: based on the data of different groups, the changing rules are obtained, and the groups with similar changing rules are preset as the control group, and the groups with opposite changing rules are preset as the observation group. When the clinical changes of the patient belong to the range of the control group, the system issues an early warning, otherwise, no early warning is issued; Step S60: Obtaining data features of susceptible individuals in historical infection cases, and dividing the obtained relevant data information into hierarchical levels; Step S70: Analyze the changing patterns of clinical manifestations at each level and the diverse disease characteristics formed, and obtain the correlation between each changing pattern and the diverse disease characteristics. At the same time, obtain the evolution pattern of clinical changing patterns at each level to susceptible data, analyze the potential risks of clinical changes at each level, and construct a risk rating table based on the potential risks.

Images