Postoperative hypoxemia risk prediction method and system

Abstract

The invention relates to a postoperative hypoxemia risk prediction method and system.The postoperative hypoxemia risk prediction method comprises the steps that patient data are obtained, and the patient data comprise the body mass index, the platelet count, the fibrinogen level and the triglyceride-glucose index; the patient data are input into a pre-trained postoperative hypoxemia risk prediction model to obtain a postoperative hypoxemia risk prediction result, and the postoperative hypoxemia risk prediction result comprises the postoperative hypoxemia occurrence probability, the 95% confidence interval and the model parameters. The method has the advantages that the preoperative data, namely BMI, PLT, FIB and TyG, can be used for predicting the postoperative hypoxemia risk of the patient, so that the patient with high postoperative hypoxemia risk can be focused after the operation; preoperative data is easy to obtain and can be obtained from routine examination of a patient, and the method is simple, rapid and convenient to operate; the speed of obtaining the postoperative hypoxemia risk prediction result is high, and the accuracy rate is high.

Description

Technical Field

[0001] The present invention relates to the field of medical nursing analysis, and in particular to a method, system, computer equipment and computer-readable storage medium for predicting the risk of postoperative hypoxemia. Background Art

[0002] The Global Obesity Observatory points out that obesity is on the rise worldwide. It is estimated that by 2030, one in five adult women and one in seven adult men will suffer from obesity, equivalent to more than 1 billion people worldwide. Bariatric surgery has obvious advantages in controlling weight and improving obesity-related diseases. As of 2018, about 61% of the major bariatric surgeries performed in the United States each year were laparoscopic sleeve gastrectomy (LSG). Currently, laparoscopic sleeve gastrectomy has also become the most important surgical method for bariatric and metabolic surgery in my country, accounting for more than 95%. However, like all other surgical operations, bariatric surgery also has risks. Postoperative pulmonary complications (PPCs) are one of the most common perioperative complications, among which postoperative hypoxemia (PH) is the most common PPC after surgery. Carlos Ferrando et al. found in a randomized controlled trial that the incidence of hypoxemia after bariatric and metabolic surgery can be as high as 80%, which is much higher than other non-cardiac surgeries in non-obese patients. High oxygen-consuming organs, such as the heart, kidneys, brain, and lungs, are prone to visceral dysfunction when the body is deprived of oxygen. Therefore, hypoxemia can lead to various complications, including brain damage, coronary artery ischemia, tachycardia, and renal dysfunction, and can even be life-threatening.

[0003] Currently, there are insufficient methods to effectively predict postoperative hypoxemia, and it is impossible to judge whether a patient will develop postoperative hypoxemia after LSG. This makes it easy for patients to suddenly develop postoperative hypoxemia, greatly increasing the risk to their lives.

[0004] Currently, no effective solution has been proposed for problems such as the inability to effectively predict postoperative hypoxemia in related technologies. Summary of the invention

[0005] The purpose of the present invention is to provide a method, system, computer device and computer-readable storage medium for predicting the risk of postoperative hypoxemia in order to solve the problems existing in the related art such as the inability to effectively predict postoperative hypoxemia.

[0006] To achieve the above object, the technical solution adopted by the present invention is:

[0007] In a first aspect, the present invention provides a method for predicting the risk of postoperative hypoxemia for obese patients undergoing laparoscopic sleeve gastrectomy, comprising:

[0008] Acquiring patient data, wherein the patient data includes body mass index, platelet count, fibrinogen level, and triglyceride-glucose index;

[0009] The patient data is input into a pre-trained postoperative hypoxemia risk prediction model to obtain a postoperative hypoxemia risk prediction result, wherein the postoperative hypoxemia risk prediction result includes the probability of postoperative hypoxemia, a 95% confidence interval, and model parameters.

[0010] In some embodiments, the postoperative hypoxemia risk prediction model is a Logistic regression equation: Y=-9.25+0.06*body weight index-0.01*platelet count+0.46*fibrinogen level+0.86*triglyceride-glucose index.

[0011] In some of the embodiments, the postoperative hypoxemia risk prediction model also includes a postoperative hypoxemia nomogram model.

[0012] In some embodiments, the method for constructing the postoperative hypoxemia risk prediction model includes:

[0013] Access patient data;

[0014] Screening the patient data to obtain risk factors, wherein the risk factors include preoperative data and intraoperative data, and the preoperative data includes patient information, preoperative blood test data, and imaging data;

[0015] Based on the patient data, construct a training set and a validation set;

[0016] The training set was analyzed using a Logistic regression equation to obtain independent risk factors, wherein the independent risk factors included body mass index, platelet count, fibrinogen level, and triglyceride-glucose index;

[0017] A postoperative hypoxemia risk prediction model is constructed based on the independent risk factors, wherein the postoperative hypoxemia risk prediction model is a Logistic regression equation: Y=-9.25+0.06*body weight index-0.01*platelet count+0.46*fibrinogen level+0.86*triglyceride-glucose index.

[0018] In some embodiments, the method for constructing the postoperative hypoxemia risk prediction model further includes:

[0019] At least one of the ROC curve, calibration curve, Hosmer-Lemeshow goodness-of-fit test, clinical decision curve, and clinical impact curve was used to verify the accuracy, calibration, and clinical utility of the postoperative hypoxemia risk prediction model.

[0020] In a second aspect, the present invention provides a postoperative hypoxemia risk prediction system, characterized in that it comprises:

[0021] an acquisition unit, configured to acquire patient data, wherein the patient data includes body mass index, platelet count, fibrinogen level, and triglyceride-glucose index;

[0022] A prediction unit is used to input the patient data into a pre-trained postoperative hypoxemia risk prediction model to obtain a postoperative hypoxemia risk prediction result, wherein the postoperative hypoxemia risk prediction result includes the probability of postoperative hypoxemia, a 95% confidence interval, and model parameters.

[0023] In some of the embodiments, it also includes:

[0024] A model building unit is used to build a risk prediction model for postoperative hypoxemia.

[0025] In some of the embodiments, the model building unit includes:

[0026] An acquisition module, used to acquire patient data;

[0027] A screening module, used to screen the patient data to obtain risk factors, wherein the risk factors include preoperative data and intraoperative data, and the preoperative data includes patient information, preoperative blood test data, and imaging data;

[0028] A set construction module, used for constructing a training set and a validation set based on the patient data;

[0029] An analysis module, used for analyzing the training set using a Logistic regression equation to obtain independent risk factors, wherein the independent risk factors include body mass index, platelet count, fibrinogen level, and triglyceride-glucose index;

[0030] A model building module is used to build a postoperative hypoxemia risk prediction model based on the independent risk factors, wherein the postoperative hypoxemia risk prediction model is Y=-9.25+0.06*body weight index-0.01*platelet count+0.46*fibrinogen level+0.86*triglyceride-glucose index.

[0031] In some of the embodiments, it also includes:

[0032] A verification unit is used to verify the accuracy, calibration and clinical utility of the postoperative hypoxemia risk prediction model by using at least one of a ROC curve, a calibration curve, a Hosmer-Lemeshow goodness-of-fit test, a clinical decision curve and a clinical impact curve.

[0033] In a third aspect, the present invention provides a computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the postoperative hypoxemia risk prediction method as described in the first aspect when executing the computer program.

[0034] In a fourth aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the method for predicting the risk of postoperative hypoxemia as described in the first aspect.

[0035] The present invention adopts the above technical solution, and has the following technical effects compared with the prior art:

[0036] The present invention provides a method, system, computer device and computer-readable storage medium for predicting the risk of postoperative hypoxemia. The method can predict the risk of postoperative hypoxemia of patients using preoperative data, namely BMI, PLT, FIB and TyG, so as to pay special attention to patients with high risk of postoperative hypoxemia after surgery. The preoperative data is easy to obtain and can be obtained from routine examinations of patients, which is simple, fast and convenient to operate. The speed of obtaining the prediction results of postoperative hypoxemia risk is fast and the accuracy is high. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 is a flow chart of a method for predicting postoperative hypoxemia risk according to an embodiment of the present invention;

[0038] Figure 2 is a schematic diagram of a postoperative hypoxemia nomogram model according to an embodiment of the present invention;

[0039] Figure 3 is a schematic diagram of a postoperative hypoxemia risk prediction system according to an embodiment of the present invention;

[0040] Figure 4 is a model research flow chart according to a specific implementation of an embodiment of the present invention;

[0041] Figure 5a is a ROC curve diagram of a training set according to an embodiment of the present invention, Figure 5b is a ROC curve diagram of a validation set according to an embodiment of the present invention;

[0042] Figure 6 is a ROC curve diagram for internal validation of a training set according to an embodiment of the present invention;

[0043] Figure 7a to Figure 7b is a calibration curve diagram of a training set according to an embodiment of the present invention;

[0044] Figure 8a is a clinical decision curve diagram of a training set according to an embodiment of the present invention, Figure 8b is a clinical decision curve diagram of a validation set according to an embodiment of the present invention;

[0045] Figure 9a is a clinical impact curve diagram of a training set according to an embodiment of the present invention, Figure 9b is a clinical impact curve diagram of a validation set according to an embodiment of the present invention.

[0046] The reference numerals are: 310, acquisition unit; 320, prediction unit; 330, construction unit; 340, verification unit. DETAILED DESCRIPTION

[0047] In order to make the purpose, technical solutions and advantages of the present application clearer, the present application is described and illustrated below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not intended to limit the present application. Based on the embodiments provided in the present application, all other embodiments obtained by ordinary technicians in the field without making creative work are within the scope of protection of the present application.

[0048] Obviously, the drawings described below are only some examples or embodiments of the present application. For ordinary technicians in this field, the present application can also be applied to other similar scenarios based on these drawings without creative work. In addition, it can also be understood that although the efforts made in this development process may be complicated and lengthy, for ordinary technicians in this field related to the content disclosed in this application, some changes in design, manufacturing or production based on the technical content disclosed in this application are just conventional technical means, and should not be understood as insufficient content disclosed in this application.

[0049] Reference to "embodiments" in this application means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present application. The appearance of the phrase in various locations in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those of ordinary skill in the art that the embodiments described in this application may be combined with other embodiments without conflict.

[0050] Unless otherwise defined, the technical terms or scientific terms involved in this application should be understood by people with ordinary skills in the technical field to which this application belongs. The words "one", "a", "a", "the" and the like involved in this application do not indicate a quantitative limitation, and may indicate the singular or plural. The terms "include", "comprise", "have" and any of their variations involved in this application are intended to cover non-exclusive inclusions; for example, a process, method, system, product or device that includes a series of steps or units (units) is not limited to the listed steps or units, but may also include steps or units that are not listed, or may also include other steps or units inherent to these processes, methods, products or devices. The words "connect", "connected", "coupled" and the like involved in this application are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. The "multiple" involved in this application refers to two or more. "And / or" describes the association relationship of associated objects, indicating that there may be three relationships, for example, "A and / or B" can mean: A exists alone, A and B exist at the same time, and B exists alone. The character " / " generally indicates that the objects before and after are in an "or" relationship. The terms "first", "second", "third", etc. involved in this application are only used to distinguish similar objects and do not represent a specific ordering of the objects.

[0051] Example 1

[0052] This embodiment relates to the method for predicting the risk of postoperative hypoxemia of the present invention.

[0053] Figure 1 FIG. 4 is a flow chart of a method for predicting postoperative hypoxemia risk according to an embodiment of the present invention. Figure 1 A method for predicting the risk of postoperative hypoxemia in obese patients undergoing laparoscopic sleeve gastrectomy is shown in Figure 1.

[0054] Step S102, obtaining patient data, wherein the patient data includes body mass index (BMI), platelet count (PLT), fibrinogen level (FIB), and triglyceride-glucose index (TyG);

[0055] Step S104: inputting the patient data into a pre-trained postoperative hypoxemia risk prediction model to obtain a postoperative hypoxemia risk prediction result, wherein the postoperative hypoxemia risk prediction result includes the probability of postoperative hypoxemia, a 95% confidence interval, and model parameters.

[0056] Among them, laparoscopic sleeve gastrectomy is a weight loss surgery for obese patients.

[0057] In step S102, TyG=Ln[fasting triglyceride (mg / dL)×fasting blood glucose (mg / dL) / 2].

[0058] In step S102, the patient data includes preoperative data, including patient information and perioperative laboratory test data.

[0059] In step S104, the postoperative hypoxemia risk prediction model is a Logistic regression equation: Y = -9.25 + 0.06 * body mass index - 0.01 * platelet count + 0.46 * fibrinogen level + 0.86 * triglyceride-glucose index, that is, Y = 9.25 + 0.06 * BMI - 0.01 * PLT + 0.46 * FIB + 0.86 * TyG.

[0060] Furthermore, the postoperative hypoxemia risk prediction model also includes Figure 2 The nomogram model for postoperative hypoxemia is shown.

[0061] In addition, the postoperative hypoxemia risk prediction method of the embodiment of the present application can be implemented by a computer device. The components of the computer device may include but are not limited to a processor and a memory storing computer program instructions.

[0062] In some embodiments, the processor 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.

[0063] In some embodiments, the memory may include a large capacity memory for data or instructions. By way of example and not limitation, the memory may include a hard disk drive (HDD), a floppy disk drive, a solid state drive (SSD), 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. Where appropriate, the memory may include a removable or non-removable (or fixed) medium. Where appropriate, the memory may be inside or outside a data processing device. In a specific embodiment, the memory is a non-volatile memory. In a specific embodiment, the memory includes a read-only memory (ROM) and a random access memory (RAM). Where appropriate, the ROM may be a mask-programmed ROM, a programmable ROM (Programmable Read-Only Memory, PROM for short), an erasable PROM (Erasable ProgrammableRead-Only Memory, EPROM for short), an electrically erasable PROM (Electrically Erasable ProgrammableRead-Only Memory, EEPROM for short), an electrically alterable ROM (Electrically Alterable Read-Only Memory, EAROM for short) or a flash memory (FLASH) or a combination of two or more of these. Under appropriate circumstances, the RAM can be a static random access memory (SRAM) or a dynamic random access memory (DRAM), wherein the DRAM can be a fast page mode dynamic random access memory (FPMDRAM), an extended data output dynamic random access memory (EDODRAM), a synchronous dynamic random access memory (SDRAM), etc.

[0064] The memory may be used to store or cache various data files required for processing and / or communication, as well as possible computer program instructions executed by the processor.

[0065] The processor implements any one of the postoperative hypoxemia risk prediction methods in the above embodiments by reading and executing computer program instructions stored in the memory.

[0066] In some embodiments, the computer device may further include a communication interface and a bus, wherein the processor, the memory, and the communication interface are connected via the bus and communicate with each other.

[0067] The communication interface is used to realize the communication between each unit, device, unit and / or device in the embodiment of the present application. The communication interface can also realize data communication with other components such as: external devices, image / data acquisition equipment, databases, external storage, and image / data processing workstations.

[0068] The bus includes hardware, software or both, and couples the components of the computer device to each other. The bus includes but is not limited to at least one of the following: data bus, address bus, control bus, expansion bus, and local bus. By way of example and not limitation, the bus may include an Accelerated Graphics Port (AGP) or other graphics bus, an Extended Industry Standard Architecture (EISA) bus, a Front Side Bus (FSB), a Hyper Transport (HT) interconnect, an Industry Standard Architecture (ISA) bus, an InfiniBand interconnect, a Low Pin Count (LPC) bus, a memory bus, a Micro Channel Architecture (MCA) bus, a Peripheral Component Interconnect (PCI) bus, a PCI-Express (PCI-X) bus, a Serial Advanced Technology Attachment (SATA) bus, a Video Electronics Standards Association Local Bus (VLB) bus, or other suitable buses or a combination of two or more of these. Where appropriate, the bus may include one or more buses. Although embodiments of the present application describe and illustrate a particular bus, the present application contemplates any suitable bus or interconnect.

[0069] The computer device can execute the postoperative hypoxemia risk prediction method in the embodiment of the present application.

[0070] In addition, in combination with the postoperative hypoxemia risk prediction method in the above embodiments, the present application embodiment can provide a computer-readable storage medium for implementation. The computer-readable storage medium stores computer program instructions; when the computer program instructions are executed by the processor, any postoperative hypoxemia risk prediction method in the above embodiments is implemented.

[0071] The technical effects of the present invention are as follows:

[0072] 1) Using preoperative data, namely BMI, PLT, FIB, and TyG, the risk of postoperative hypoxemia can be predicted, so that patients with a high risk of postoperative hypoxemia can be given special attention after surgery;

[0073] 2) Preoperative data is easy to obtain and can be obtained from the patient's routine examination, which is simple, fast and convenient to operate;

[0074] 3) The speed of obtaining the prediction results of postoperative hypoxemia risk is fast and the accuracy is high.

[0075] Example 2

[0076] This embodiment relates to a method for constructing a postoperative hypoxemia risk prediction model of the present invention.

[0077] In an illustrative embodiment of the present invention, a method for constructing a postoperative hypoxemia risk prediction model includes:

[0078] Step S202: obtaining patient data;

[0079] Step S204, screening the patient data to obtain risk factors, wherein the risk factors include preoperative data and intraoperative data, and the preoperative data includes patient information, preoperative blood test data, and imaging data;

[0080] Step S206: construct a training set and a validation set based on the patient data;

[0081] Step S208, using a Logistic regression equation to analyze the training set to obtain independent risk factors, wherein the independent risk factors include body mass index, platelet count, fibrinogen level, and triglyceride-glucose index;

[0082] Step S210: construct a postoperative hypoxemia risk prediction model based on independent risk factors, wherein the postoperative hypoxemia risk prediction model is a Logistic regression equation: Y=-9.25+0.06*body weight index-0.01*platelet count+0.46*fibrinogen level+0.86*triglyceride-glucose index.

[0083] In step S202 , the patient data is patient data of an obese patient who has undergone laparoscopic sleeve gastrectomy.

[0084] In step S204 , the patient information includes age, gender, body mass index (BMI), smoking history, hypertension, diabetes, asthma, obstructive sleep apnea, and American Society of Anesthesiologists (ASA) classification.

[0085] In step S204, the preoperative blood test data (i.e., perioperative laboratory tests) include preoperative arterial blood pH, arterial partial pressure of oxygen (PaO2), arterial partial pressure of carbon dioxide (PaCO2), arterial lactate (oxygen saturation in the arterial blood, SaO2), arterial oxygen saturation (oxygen saturation in the arterial blood, SaO2), white blood cell count (WBC), red blood cell count (RBC), hemoglobin (Hb), platelet count (PLT), neutrophil count (N), lymphocyte count (L), monocyte count (M), erythrocyte pressure (HCT), erythrocyte sedimentation rate (ESR), and erythrocyte sedimentation rate (ESR). estradiol (ESR), DD dimer, fibrinogen (FIB), albumin (ALB), globulin (GLB), total billirubin (TBIL), direct bilirubin (DBIL), alanine aminotransferase (ALT), aspartate aminotransferase (AST), serum creatinine (Scr), uric acid (UA), total cholesterol (TC), triglycerides (TG), high-density lipoprotein (HDL), low-density lipoprotein (LDL), free fatty acids (FFA), fasting plasma glucose (FPG), glycosylated hemoglobin (HbA1c), C-reactive protein (CRP), interleukin-6 (IL6), tumor necrosis factor (TNFα), triglyceride-glucose (TyG) index (TyG index calculation formula is Ln [fasting triglyceride (mg / dL) × fasting blood glucose (mg / dL) / 2]), the lowest arterial oxygen partial pressure within 24 hours after surgery, and the oxygen flow at the lowest arterial oxygen partial pressure.

[0086] In step S204, the imaging data includes left ventricular ejection fraction (LVEF) measured by echocardiography.

[0087] In step S204, the intraoperative data includes the operation time and the intraoperative blood loss.

[0088] In step S210, it also includes:

[0089] Based on the Logistic regression equation: Y = -9.25 + 0.06*BMI-0.01*PLT + 0.46*FIB + 0.86*TyG, a nomogram model was drawn.

[0090] Furthermore, at least one of the ROC curve, calibration curve, Hosmer-Lemeshow goodness-of-fit test, clinical decision curve and clinical impact curve was used to verify the accuracy, calibration and clinical utility of the postoperative hypoxemia risk prediction model.

[0091] The technical effects of the present invention are as follows:

[0092] 1) Based on the rehabilitation management of laparoscopic sleeve gastrectomy, a risk prediction model for postoperative hypoxemia was constructed using preoperative data;

[0093] 2) The internal and external validity of the model were systematically verified using multiple indicators, and the postoperative hypoxemia risk prediction model had good predictive ability;

[0094] 3) The postoperative hypoxemia risk prediction model constructed by the present invention helps to improve the overall care of patients, better perform postoperative intervention on patients, and ensure clinical patient safety.

[0095] Example 3

[0096] This embodiment relates to the postoperative hypoxemia risk prediction system of the present invention.

[0097] Figure 3is a framework diagram of a postoperative hypoxemia risk prediction system according to an embodiment of the present invention. Figure 4 As shown, a postoperative hypoxemia risk prediction system includes an acquisition unit 310 and a prediction unit 320. The acquisition unit 310 is used to acquire patient data, wherein the patient data includes body mass index, platelet count, fibrinogen level, and triglyceride-glucose index; the prediction unit 320 is used to input the patient data into a pre-trained postoperative hypoxemia risk prediction model to obtain a postoperative hypoxemia risk prediction result, wherein the postoperative hypoxemia risk prediction result includes the probability of postoperative hypoxemia, 95% confidence interval, and model parameters.

[0098] Furthermore, the postoperative hypoxemia risk prediction system further includes a model building unit 330. The model building unit 330 is used to build a postoperative hypoxemia risk prediction model.

[0099] Specifically, the model building unit 330 includes an acquisition module, a screening module, a set building module, an analysis module and a model building module. The acquisition module is used to acquire patient data; the screening module is used to screen the patient data to obtain risk factors, wherein the risk factors include preoperative data and intraoperative data, and the preoperative data include patient information, preoperative blood test data, and imaging data; the set building module is used to build a training set and a validation set based on the patient data; the analysis module is used to analyze the training set using a Logistic regression equation to obtain independent risk factors, wherein the independent risk factors include body mass index, platelet count, fibrinogen level, and triglyceride-glucose index; the model building module is used to build a postoperative hypoxemia risk prediction model based on the independent risk factors, wherein the postoperative hypoxemia risk prediction model is Y=-9.25+0.06*body mass index-0.01*platelet count+0.46*fibrinogen level+0.86*triglyceride-glucose index.

[0100] Furthermore, the postoperative hypoxemia risk prediction system further includes a verification unit 340. The verification unit 340 is used to verify the accuracy, calibration and clinical utility of the postoperative hypoxemia risk prediction model by using at least one of a ROC curve, a calibration curve, a Hosmer-Lemeshow goodness of fit test, a clinical decision curve and a clinical impact curve.

[0101] The technical effects of this embodiment are basically the same as those of Embodiments 1 to 2, and will not be described in detail here.

[0102] Example 4

[0103] This embodiment is a specific embodiment of the present invention.

[0104] Figure 4 This is a model research flow chart of a specific embodiment of the present invention. Figure 4 As shown in the figure, the method for constructing a postoperative hypoxemia risk prediction model includes:

[0105] (I) Constructing a training set

[0106] Data of patients who underwent LSG in Shanghai Tenth People's Hospital were collected;

[0107] The above patient data were screened (screened according to the exclusion criteria, see Figure 4 );

[0108] Construct a training set using the screened patient data;

[0109] The training set was screened by multi-factor backward stepwise regression method to obtain variables;

[0110] Based on the variables, a risk prediction model for postoperative hypoxemia (including a nomogram model for postoperative hypoxemia) was constructed.

[0111] (II) Constructing a validation set

[0112] Data of patients who underwent LSG in a hospital in Beijing and a hospital in Chengdu were collected;

[0113] The above patient data were screened (screened according to the exclusion criteria, see Figure 4 );

[0114] The screened patient data was used to construct a validation set;

[0115] (III) Verification

[0116] The data of the validation set were input into the postoperative hypoxemia risk prediction model for validation.

[0117] The specific steps are as follows:

[0118] 1. Research subjects

[0119] This was a retrospective cohort study, retrospectively investigating obese patients who underwent laparoscopic sleeve gastrectomy (LSG) in Shanghai Tenth People's Hospital from January 2018 to December 2023.

[0120] The inclusion criteria include:

[0121] (1) Body mass index (BMI) ≥ 27.5 kg / m 2 patients;

[0122] (2) patients aged between 18 and 65 years;

[0123] (3) Patients undergoing LSG surgery.

[0124] Among them, the exclusion criteria include:

[0125] (1) Patients who have undergone bariatric surgery in the past;

[0126] (2) Patients with malignant tumors or immune diseases;

[0127] (3) patients who died during surgery or within 24 hours after surgery;

[0128] (4) patients with missing arterial blood gas and oxygen flow data within 24 hours after surgery;

[0129] (5) Patients whose variable values ​​are missing for more than 50% of the total variables.

[0130] A total of 436 eligible patients who underwent LSG surgery were screened for this study. Among them, 13 patients had undergone bariatric surgery before, 24 patients had malignant tumors or immune diseases, 3 patients died during surgery or within 24 hours after surgery, 172 patients had missing arterial blood gas and oxygen flow data within 24 hours after surgery, and 29 patients with missing variable values ​​exceeding 50% of the total variables were excluded. Therefore, 195 patients finally met the inclusion and exclusion criteria.

[0131] Among them, a total of 132 patients among these 195 patients suffered from postoperative hypoxemia, i.e., the incidence rate was 67.69%.

[0132] 2. Data Collection

[0133] This study included more than 40 perioperative data, including:

[0134] (1) Patient information: age, gender, body mass index (BMI), smoking history, hypertension, diabetes, asthma, obstructive sleep apnea, and American Society of Anesthesiologists (ASA) classification;

[0135] (2) Perioperative laboratory tests: preoperative arterial pH, arterial partial pressure of oxygen (PaO2), arterial partial pressure of carbon dioxide (PaCO2), arterial lactate (oxygen saturation in the arterial blood, SaO2), arterial oxygen saturation (oxygen saturation in the arterial blood, SaO2), white blood cell count (WBC), red blood cell count (RBC), hemoglobin (Hb), platelet count (PLT), neutrophil count (N), lymphocyte count (L), monocyte count (M), erythrocyte pressure (HCT), erythrocyte sedimentation rate (ESR), and erythrocyte saturation (ESR). estradiol (ESR), DD dimer, fibrinogen (FIB), albumin (ALB), globulin (GLB), total bilirubin (TBIL), direct bilirubin (DBIL), alanine aminotransferase (ALT), aspartate aminotransferase (AST), serum creatinine (Scr), uric acid (UA), total cholesterol (TC), triglycerides (TG), high-density lipoprotein (HDL), low-density lipoprotein (LDL), free fatty acids (FFA), fasting plasma glucose (FPG), glycosylated hemoglobin (HbA1c), c-reactive protein (CRP),CRP), interleukin-6 (IL6), tumor necrosis factor (TNFα), triglyceride-glucose (TyG) index (TyG index calculation formula is Ln [fasting triglyceride (mg / dL) × fasting blood glucose (mg / dL) / 2]), the lowest arterial oxygen partial pressure within 24 hours after surgery, and the oxygen flow at the lowest arterial oxygen partial pressure;

[0136] (3) Imaging data: left ventricular ejection fraction (LVEF) measured by echocardiography,

[0137] (4) Intraoperative data: operation time and intraoperative blood loss.

[0138] 3. Diagnostic criteria for postoperative hypoxemia (PH)

[0139] According to the diagnostic criteria for acute respiratory distress syndrome updated at the 2012 Berlin Conference, hypoxemia was defined as arterial oxygen partial pressure / inspired oxygen fraction (PaO2 / FiO2) ≤ 300 mmHg within 24 hours after surgery.

[0140] The lowest oxygenation index within 24 hours after surgery was calculated to determine whether the patients had hypoxemia after surgery, and the patients were divided into a hypoxemia group (H group): PaO2 / FiO2≤300 mmHg and a non-hypoxemia group (NH group): PaO2 / FiO2>300 mmHg.

[0141] 4. Baseline data analysis

[0142] The variables of the patients in the NH and H groups were analyzed at baseline. Two independent sample t-tests were used to compare the continuous variables with normal distribution, and the Mann-Whitney test was used for the continuous variables with non-normal distribution. If the data of the two groups were categorical variables, the chi-square test or Fisher's exact test was used to compare the two groups. The baseline data of the patients were used as candidate parameters for model development.

[0143] The modeling process of this study was implemented using R language (version 4.2.1, https: / / www.R-project.org).

[0144] 5. Modeling variable screening

[0145] All baseline data were included in univariate logistic regression, and statistically significant variables (P<0.05) were screened out through univariate logistic regression analysis and included in the multivariate logistic regression equation. The reverse stepwise regression method based on the Akaike information criterion (AIC) was used to determine the variables retained in the prediction model.

[0146] 6. Model construction

[0147] The obtained modeling variables were used to establish the nomogram model.

[0148] 7. Model Validation

[0149] After the nomogram model was established, the model was validated using internal and external validation groups.

[0150] The model was internally validated using 500 bootstrap resampling methods in the training set. Obese patients who underwent laparoscopic sleeve gastrectomy in a hospital in Beijing and a hospital in Chengdu from January 2021 to December 2023 were included as the external validation set for external validation. The discriminatory ability of the nomogram model was evaluated by calculating the receiver operating characteristic (ROC) curve and the area under the curve (AUC). The calibration curve and the Hosmer-Lemeshow goodness-of-fit test were used to compare the relationship and consistency between the predicted probability and the actual results. The clinical practicality of the nomogram was evaluated by clinical decision curve analysis (DCA) and drawing the clinical impact curve (CIC).

[0151] 8. Results

[0152] 8.1 Baseline data

[0153] A total of 195 patients from Shanghai Tenth People's Hospital were included in the training set, of which 103 patients (52.82%) were male, with a median age of 30 years (IQR 24-35) and a median BMI of 42.14 kg / m 2 (IQR 38.76-48.46), and the incidence of hypoxemia was 67.69% (Table 1).

[0154] The external validation cohort included 105 patients from a hospital in Beijing and a hospital in Chengdu, 67 patients (63.81%) were male, the median age was 33 years (IQR 28-41), the average BMI was 46.89 kg / m2 (SD 10.44), and the incidence of hypoxemia was 75.24% (Table 2). We compared the baseline data between the training set and the validation set.

[0155] Compared with the training set, the patients in the validation set had greater age, BMI, red blood cell count, hemoglobin, neutrophil count, glycated hemoglobin, total bilirubin, indirect bilirubin, and fasting glucose, but smaller intraoperative blood loss, preoperative SaO2, platelets, lymphocyte count, albumin, and D-dimer. There were more smoking patients, hypertension patients, diabetes patients, and obstructive sleep apnea patients in the validation set. No significant differences were observed in other clinical characteristics (Table 3).

[0156] Table 1 Baseline characteristics of patients in the training set

[0157]

[0158]

[0159]

[0160] Table 2 Baseline characteristics of patients in the validation set

[0161]

[0162]

[0163]

[0164]

[0165] Table 3 Comparison of baseline characteristics of patients in the training set and validation set

[0166]

[0167]

[0168]

[0169] 8.2 Variable Screening

[0170] In the training set, univariate logistic regression analysis showed that five variables were statistically significant, namely, preoperative BMI, platelets, glycosylated hemoglobin, fibrinogen, and TyG index (p < 0.05) (Table 4). Whether hypoxemia occurred in obese patients within 24 hours after laparoscopic sleeve gastrectomy was the dependent variable, and the above five variables were considered as independent variables.

[0171] Multivariate logistic regression analysis was performed using the reverse stepwise method, and BMI (OR=1.06, 95%CI=1.01, 1.12, p=0.035), platelets (OR=0.99, 95%CI=0.99, 1.00, p=0.011), fibrinogen (OR=1.59, 95%CI=1.02, 2.57, p=0.050) and TyG index (OR=2.37, 95%CI=1.42, 4.16, p=0.002) were retained in the final prediction model (Table 5).

[0172] Table 4 Univariate logistic regression analysis of the training set

[0173]

[0174]

[0175]

[0176] Table 5 Multivariate logistic regression analysis of the training set

[0177]

[0178] 8.3 Constructing a Nomogram

[0179] Based on the logistic regression coefficient and constant term, a prediction model was constructed to predict the probability of hypoxemia in obese patients within 24 hours after sleeve gastrectomy.

[0180] The logistic regression equation is as follows: Logit(P)=-9.25+0.06×BMI-0.01×PLT+0.46×FIB+0.86×TyG.

[0181] A nomogram was established based on the results of multivariate logistic regression analysis ( Figure 2). To predict the probability of hypoxemia in obese patients within 24 hours after sleeve gastrectomy, draw a line perpendicular to the corresponding indicator to the point scale on the nomogram based on each patient's BMI, preoperative PLT, FIB, and calculated TyG index, and add the points of each indicator. Next, mark the sum on the total score line and draw a line perpendicular to the risk axis, which is the probability of hypoxemia in obese patients within 24 hours after weight loss surgery. Users can submit 4 features (BMI, PLT, FIB, and TyG index) for calculation via a computer or mobile phone. After calculating the sample output, the probability of postoperative hypoxemia, the 95% confidence interval, and the parameters of the model will be displayed.

[0182] 8.4 Model Validation

[0183] The discrimination of the nomogram model is measured by the ROC curve in the training set and the validation set ( Figure 5a , Figure 5b The AUC of the training set was 0.733 (95% CI = 0.655, 0.811), and the AUC of the internal validation using the bootstrap repeated sampling method was 0.757 (95% CI = 0.686, 0.828) ( Figure 6 ), while the AUC of the validation set was 0.796 (95% CI = 0.700, 0.892). These results indicate that the nomogram model exhibits strong discrimination ability and predictive value. In the training set, we calculated the maximum Youden index to determine that the optimal critical value of the risk score was 0.688 (95% CI = 0.674, 0.746). Using the critical value as the threshold, obese patients undergoing bariatric surgery were divided into high-risk and low-risk groups. The calibration curves of the nomogram model in the training and validation sets found that the curves of the prediction model and the calibration curves after the bootstrap repeated sampling method were close to the ideal diagonal line ( Figure 7a , Figure 7b ). The Hosmer-Lemeshow goodness-of-fit test showed that the χ2 value of the development group was 8.38, the P value was 0.397, and the χ2 value of the validation group was 6.22, the P value was 0.623. The results of the study showed that the nomogram showed good consistency between the predicted probability and the actual probability, which was helpful to accurately predict the risk of postoperative hypoxemia. The clinical decision curves of the training set and the validation set showed that when the threshold probability was between 42% and 84% and between 40% and 100%, respectively, using the nomogram to predict the probability of postoperative hypoxemia and intervening would obtain more net benefits than intervening in all patients or not intervening in all patients ( Figure 8a , Figure 8b ). The clinical impact curves of both the training set and the validation set indicated that when the threshold probability was greater than 40%, the number of predicted postoperative hypoxemia was close to the actual number of positive cases ( Figure 9a , Figure 9b). These findings support the strong clinical utility of the nomogram in predicting hypoxemia after bariatric surgery.

[0184] in conclusion:

[0185] For the present invention, this study is the first to explore the risk factors for hypoxemia after sleeve gastrectomy in obese patients and to construct and validate a prediction model. Secondly, this study is the first to link the TyG index with postoperative hypoxemia. The results of this study broaden the clinical application of the TyG index and further reveal the possible mechanism of postoperative hypoxemia.

[0186] In addition, although the postoperative hypoxemia risk prediction method of the present invention can be used to assess the risk of postoperative hypoxemia in patients, even if the assessment result is a high risk, it will not lead to patients not undergoing surgery or postponing surgery. Instead, patients with high risk will be given special attention after surgery, and timely responses can be made when high-risk patients develop hypoxemia.

[0187] The above description is only a preferred embodiment of the present invention, and does not limit the implementation mode and protection scope of the present invention. For those skilled in the art, it should be aware that all solutions obtained by equivalent substitutions and obvious changes made using the description and illustrations of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for predicting the risk of postoperative hypoxemia for obese patients undergoing laparoscopic sleeve gastrectomy, characterized in that: include: Acquiring patient data, wherein the patient data includes body mass index, platelet count, fibrinogen level, and triglyceride-glucose index; The patient data is input into a pre-trained postoperative hypoxemia risk prediction model to obtain a postoperative hypoxemia risk prediction result, wherein the postoperative hypoxemia risk prediction result includes the probability of postoperative hypoxemia, a 95% confidence interval, and model parameters.

2. The method for predicting postoperative hypoxemia risk according to claim 1, characterized in that: The postoperative hypoxemia risk prediction model is a Logistic regression equation: Y=-9.25+0..06*body weight index-0.01*platelet count+0.46*fibrinogen level+0.86*triglyceride-glucose index.

3. The method for predicting postoperative hypoxemia risk according to claim 2, characterized in that: The postoperative hypoxemia risk prediction model also includes a postoperative hypoxemia nomogram model.

4. The method for predicting postoperative hypoxemia risk according to claims 1 to 3, characterized in that: The method for constructing the postoperative hypoxemia risk prediction model comprises: Access patient data; Screening the patient data to obtain risk factors, wherein the risk factors include preoperative data and intraoperative data, and the preoperative data includes patient information, preoperative blood test data, and imaging data; Based on the patient data, construct a training set and a validation set; The training set was analyzed using a Logistic regression equation to obtain independent risk factors, wherein the independent risk factors included body mass index, platelet count, fibrinogen level, and triglyceride-glucose index; A postoperative hypoxemia risk prediction model is constructed based on the independent risk factors, wherein the postoperative hypoxemia risk prediction model is a Logistic regression equation: Y=-9.25+0.06*body weight index-0.01*platelet count+0.46*fibrinogen level+0.86*triglyceride-glucose index.

5. According to the method for predicting the risk of postoperative hypoxemia according to claim 4, the method for constructing the postoperative hypoxemia risk prediction model further comprises: At least one of the ROC curve, calibration curve, Hosmer-Lemeshow goodness-of-fit test, clinical decision curve, and clinical impact curve was used to verify the accuracy, calibration, and clinical utility of the postoperative hypoxemia risk prediction model.

6. A postoperative hypoxemia risk prediction system, characterized in that: include: an acquisition unit, configured to acquire patient data, wherein the patient data includes body mass index, platelet count, fibrinogen level, and triglyceride-glucose index; A prediction unit is used to input the patient data into a pre-trained postoperative hypoxemia risk prediction model to obtain a postoperative hypoxemia risk prediction result, wherein the postoperative hypoxemia risk prediction result includes the probability of postoperative hypoxemia, a 95% confidence interval, and model parameters.

7. The postoperative hypoxemia risk prediction system according to claim 8, characterized in that: Also includes: A model building unit is used to build a risk prediction model for postoperative hypoxemia.

8. The postoperative hypoxemia risk prediction system according to claim 7, characterized in that: The model building unit comprises: An acquisition module, used to acquire patient data; A screening module, used to screen the patient data to obtain risk factors, wherein the risk factors include preoperative data and intraoperative data, and the preoperative data includes patient information, preoperative blood test data, and imaging data; A set construction module, used for constructing a training set and a validation set based on the patient data; An analysis module, used for analyzing the training set using a Logistic regression equation to obtain independent risk factors, wherein the independent risk factors include body mass index, platelet count, fibrinogen level, and triglyceride-glucose index; A model building module is used to build a postoperative hypoxemia risk prediction model based on the independent risk factors, wherein the postoperative hypoxemia risk prediction model is Y=-9.25+0.06*body weight index-0.01*platelet count+0.46*fibrinogen level+0.86*triglyceride-glucose index.

9. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the method for predicting the risk of postoperative hypoxemia as claimed in any one of claims 1 to 5 is implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the postoperative hypoxemia risk prediction method as claimed in any one of claims 1 to 5 is implemented.

Images