Prediction method and system for potassium ion concentration grade in blood

Abstract

The invention discloses a method and a system for predicting the concentration grade of potassium ions in blood. The method comprises the following steps: firstly, collecting the concentration grade of the potassium ions and clinical data of a lumbar postoperative patient; the potassium ion concentration grade comprises normal potassium ion concentration and low potassium ion concentration; performing single-factor analysis on the clinical data to obtain associated clinical data with high correlation; carrying out LASSO regression analysis on the associated clinical data to obtain key clinical data; and finally, according to the key clinical data, establishing a column diagram for predicting the probability of low potassium ion concentration. According to the method, multiple clinical characteristic expressions can be comprehensively evaluated, so that the probability that the potassium ion concentration in the blood is relatively low is accurately predicted.

Description

Technical Field

[0001] The present invention relates to potassium ion concentration prediction, in particular to a method and system for predicting potassium ion concentration levels in blood. Background Art

[0002] Based on the concept of perioperative management, medical staff are required to conduct more accurate assessment and monitoring before, during and after lumbar surgery. Potassium, as one of the essential macroelements in the human body, plays a vital role in maintaining cell membrane function, regulating osmotic pressure, maintaining muscle excitability and many other aspects. When the blood potassium concentration is <3.5mmol / L, it is considered to be low, and the patient is at higher risk of myocardial injury and prognosis. However, the cause of low blood potassium concentration in patients undergoing lumbar surgery is usually unknown, so it is difficult to accurately and effectively assess blood potassium concentration based on baseline data and clinical examinations. Summary of the invention

[0003] Purpose of the invention: The purpose of the present invention is to provide a method and system for predicting the level of potassium ion concentration in the blood, which can screen key features based on clinical data and accurately predict the situation of low potassium ion concentration in the blood.

[0004] Technical solution: The method for predicting the potassium ion concentration level in blood of the present invention comprises the following steps:

[0005] (1) To collect potassium ion concentration levels and clinical data of patients after lumbar spine surgery; potassium ion concentration levels include normal potassium ion concentration and low potassium ion concentration;

[0006] (2) performing univariate analysis on the clinical data to obtain correlated clinical data with high correlation;

[0007] (3) performing LASSO regression analysis on the associated clinical data to obtain key clinical data;

[0008] (4) establishing a nomogram based on the key clinical data to predict the probability of low potassium ion concentration.

[0009] Furthermore, the clinical data in step (1) include gender, age, BMI, medical history, and laboratory test results.

[0010] Furthermore, in step (2), the clinical data are subjected to univariate analysis, and clinical data with a P value less than 0.05 are screened as associated clinical data.

[0011] Furthermore, the key clinical data in step (3) include age, disease type, diabetes, urea nitrogen, preoperative potassium and drainage volume, and the disease types include lumbar disc herniation, lumbar stenosis and lumbar disc herniation combined with lumbar stenosis.

[0012] The system for predicting the potassium ion concentration level in blood of the present invention comprises:

[0013] A clinical data collection unit, used to collect potassium ion concentration levels and clinical data of patients after lumbar spine surgery; the potassium ion concentration levels include normal potassium ion concentration and low potassium ion concentration;

[0014] A related clinical data screening unit, used for performing univariate analysis on the clinical data to obtain related clinical data with high correlation;

[0015] A key clinical data screening unit, used for performing LASSO regression analysis on the associated clinical data to obtain key clinical data;

[0016] The potassium ion concentration prediction unit establishes a nomogram based on the key clinical data to predict the probability of low potassium ion concentration.

[0017] Furthermore, the clinical data in the clinical data collection unit includes gender, age, BMI, medical history, and laboratory test results.

[0018] Furthermore, in the associated clinical data screening unit, a univariate analysis is performed on the clinical data, and clinical data with a P value less than 0.05 are screened as associated clinical data.

[0019] Furthermore, the key clinical data in the key clinical data screening unit include age, disease type, diabetes, urea nitrogen, preoperative potassium and drainage volume, and the disease types include lumbar disc herniation, lumbar stenosis and lumbar disc herniation combined with lumbar stenosis.

[0020] The electronic device described in the present invention includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the computer program is loaded into the processor, the method for predicting the potassium ion concentration level in the blood is implemented.

[0021] The computer-readable storage medium of the present invention stores a computer program, and when the computer program is executed by a processor, the method for predicting the potassium ion concentration level in the blood is implemented.

[0022] Beneficial effect: Compared with the prior art, the advantages of the present invention are: the present invention comprehensively evaluates the baseline data and clinical characteristics of patients after lumbar surgery, obtains factors closely related to the blood potassium concentration level through multiple analyses and screenings, and then accurately predicts the probability of low potassium ion concentration in the blood, avoiding the problem of misjudgment caused by one-sided information. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 is a regression coefficient path diagram of an embodiment of the present invention;

[0024] Figure 2 A cross-validation curve diagram of an embodiment of the present invention;

[0025] Figure 3 is a nomogram of an embodiment of the present invention;

[0026] Figure 4 is a ROC curve diagram of an embodiment of the present invention;

[0027] Figure 5 is a calibration curve diagram of an embodiment of the present invention;

[0028] Figure 6 This is a DCA decision analysis curve diagram of an embodiment of the present invention. DETAILED DESCRIPTION

[0029] The technical solution of the present invention is further described below in conjunction with the accompanying drawings. The method for predicting the potassium ion concentration in blood of the present invention comprises the following steps.

[0030] Step 1: Collect potassium ion concentration levels and clinical data of patients after lumbar surgery from historical cases, including gender, age, BMI, medical history, laboratory tests, etc.

[0031] Step 2, perform univariate analysis on the clinical data, and screen the associated clinical data associated with the potassium ion concentration level. This embodiment performs univariate analysis on 46 clinical data, first checking whether they conform to the normal distribution. The data conforming to the normal distribution use the chi-square test, expressed as mean ± standard deviation (M ± SD); the count data conforming to the normal distribution use the independent sample T test, expressed as count and percentage [N (%)]; calculate the P value, in this study, P < 0.05 (bilateral) is statistically significant. The results are shown in Table 1.

[0032] Table 1 Results of univariate analysis of clinical data

[0033]

[0034]

[0035] From the results in Table 1, we can see that age (P < 0.001), preoperative RBC (P = 0.002), preoperative HGB (P = 0.014), preoperative urea nitrogen (P = 0.01), preoperative potassium (P < 0.001), blood transfusion volume (P = 0.018), intraoperative bleeding (P = 0.002), drainage volume (P < 0.001), disease type (P = 0.003), hypertension (P = 0.05), diabetes (P = 0.005), cardiovascular and cerebrovascular diseases (P = 0.008), surgical approach (P = 0.001), ASA grade (P = 0.11), osteoporosis (P = 0.036) and drainage volume (P = 0.005) are significant related factors for the potassium ion concentration level after lumbar surgery.

[0036] Step 3: For the 16 related clinical data obtained in step 2, a penalty function model is constructed using the LASSO technology, the coefficients are adjusted to zero, and 6 key clinical data are screened, as shown in Table 2.

[0037] Table 2 LASSO construction variables

[0038] variable LASSO coefficient Regression coefficient (β) age 0.0393027106 0.0625 Disease Type 0.1908623212 0.5415 diabetes 0.0143631221 0.5360 Urea nitrogen -0.0619111153 -0.2253 Preoperative potassium -0.3005515317 -0.8870 Traffic 0.0001027155 0.0002

[0039] Figure 1 and Figure 2 They are the regression coefficient path diagram and the cross-validation curve diagram. By adjusting the lambda value to reduce the coefficient to zero, the 46 included variables are gradually compressed, and finally the 6 best variables are screened out.

[0040] Step 4, establishing a nomogram for predicting potassium ion concentration levels.

[0041] Use the six key clinical data obtained in step 3 to establish a nomogram, such as Figure 3 As shown, by adding the corresponding scores of the six key clinical data on the vertical lines, and then matching the total score with the scale below, the probability of low potassium ion concentration is obtained. For example, a 65-year-old patient with lumbar disc herniation and stenosis, diabetes, uric acid nitrogen level of 12, preoperative potassium level of 5, and postoperative drainage volume of 2000. The corresponding scores of this patient are 70, 23, 12, 20, 40, and 10, and the total score is 175, corresponding to the probability of low potassium ion concentration of about 0.2.

[0042] Figure 4 The ROC curve analysis is shown, in which the AUC is 0.786 (95% CI 0.734-0.837), which shows that it has a high accuracy in predicting the probability of low potassium ion concentration. Figure 5 The calibration curve shows good agreement between the predicted and observed results for low potassium concentrations. Figure 6The DCA decision analysis curve shows a high net benefit in predicting the probability of low potassium concentration.

[0043] The system for predicting the potassium ion concentration level in blood of the present invention comprises:

[0044] A clinical data collection unit, used to collect potassium ion concentration levels and clinical data of patients after lumbar spine surgery; the potassium ion concentration levels include normal potassium ion concentration and low potassium ion concentration;

[0045] A related clinical data screening unit, used for performing univariate analysis on the clinical data to obtain related clinical data with high correlation;

[0046] A key clinical data screening unit, used for performing LASSO regression analysis on the associated clinical data to obtain key clinical data;

[0047] The potassium ion concentration prediction unit establishes a nomogram based on the key clinical data to predict the probability of low potassium ion concentration.

[0048] The electronic device described in the present invention includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the computer program is loaded into the processor, the method for predicting the potassium ion concentration level in the blood is implemented.

[0049] The computer-readable storage medium of the present invention stores a computer program, and when the computer program is executed by a processor, the method for predicting the potassium ion concentration level in the blood is implemented.

[0050] The computer-readable storage media may include RAM, ROM, EEPROM, CD-ROM or other optical disk storage, magnetic disk storage or other magnetic storage devices, flash memory, or any other medium that can be used to store program code in the form of instructions or data structures and that can be accessed by a computer.

[0051] The processor is used to execute the computer program stored in the memory to implement each step of the method involved in the above embodiment.

Claims

1. A method for predicting the potassium ion concentration level in blood, characterized in that: The steps include: (1) To collect potassium ion concentration levels and clinical data of patients after lumbar spine surgery; potassium ion concentration levels include normal potassium ion concentration and low potassium ion concentration; (2) performing univariate analysis on the clinical data to obtain correlated clinical data with high correlation; (3) performing LASSO regression analysis on the associated clinical data to obtain key clinical data; (4) establishing a nomogram based on the key clinical data to predict the probability of low potassium ion concentration.

2. The method for predicting the potassium ion concentration level in blood according to claim 1, characterized in that: The clinical data in step (1) include gender, age, BMI, medical history, and laboratory test results.

3. The method for predicting the potassium ion concentration level in blood according to claim 1, characterized in that: In step (2), the clinical data are subjected to univariate analysis, and clinical data with a P value less than 0.05 are selected as associated clinical data.

4. The method for predicting the potassium ion concentration level in blood according to claim 1, characterized in that: The key clinical data described in step (3) include age, disease type, diabetes, urea nitrogen, preoperative potassium and drainage volume, and the disease types include lumbar disc herniation, lumbar stenosis and lumbar disc herniation combined with lumbar stenosis.

5. A system for predicting the potassium ion concentration level in blood, characterized in that: include: A clinical data collection unit, used to collect potassium ion concentration levels and clinical data of patients after lumbar spine surgery; the potassium ion concentration levels include normal potassium ion concentration and low potassium ion concentration; A related clinical data screening unit, used for performing univariate analysis on the clinical data to obtain related clinical data with high correlation; A key clinical data screening unit, used for performing LASSO regression analysis on the associated clinical data to obtain key clinical data; The potassium ion concentration prediction unit establishes a nomogram based on the key clinical data to predict the probability of low potassium ion concentration.

6. The method for predicting the potassium ion concentration level in blood according to claim 5, characterized in that: The clinical data in the clinical data collection unit include gender, age, BMI, medical history, and laboratory test results.

7. The method for predicting the potassium ion concentration level in blood according to claim 5, characterized in that: In the associated clinical data screening unit, a univariate analysis is performed on the clinical data, and clinical data with a P value less than 0.05 are screened as associated clinical data.

8. The method for predicting the potassium ion concentration level in blood according to claim 5, characterized in that: The key clinical data described in the key clinical data screening unit include age, disease type, diabetes, urea nitrogen, preoperative potassium and drainage volume. The disease types include lumbar disc herniation, lumbar stenosis and lumbar disc herniation combined with lumbar stenosis.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the computer program is loaded into a processor, the method for predicting the potassium ion concentration level in the blood according to any one of claims 1 to 4 is implemented.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the method for predicting the potassium ion concentration level in the blood according to any one of claims 1 to 4 is implemented.

Images