Neurosurgery postoperative CT bleeding risk assessment method, assessment model and establishment method thereof

By establishing a neurosurgery postoperative CT bleeding risk assessment model based on R language, using preoperative, intraoperative and postoperative data screening factors, the problem of strong occultness of postoperative bleeding was solved, and the accuracy of postoperative bleeding risk was achieved Predictions reduce the frequency and cost of CT examinations and improve patient safety.

CN119008019APending Publication Date: 2024-11-22AFFILIATED HUSN HOSPITAL OF FUDAN UNIV
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411023823.6
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-29
Publication Date
2024-11-22

AI Technical Summary

Technical Problem

Neurosurgery is highly occult, and the existing CT examination increases the patient's radiation exposure and medical costs, and early intervention windows are easily missed, making it difficult to accurately judge the risk of postoperative bleeding.

Method used

Establish a risk assessment model for postoperative CT bleeding in neurosurgery based on R language. By sorting out the original data of preoperative, intraoperative and postoperative, univariate and multivariate logistics regression methods are used to screen statistically significant dependent variables to form an evaluation model. The probability of bleeding is calculated including the type of disease, gender, hospitalization time, surgical site and duration of surgery.

Benefits of technology

It provides a more accurate basis to determine whether early CT examination is needed, reduces radiation exposure and costs, improves the accuracy and reliability of bleeding risks, and ensures patient safety.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119008019A_ABST
    Figure CN119008019A_ABST
Patent Text Reader

Abstract

The invention provides a neurosurgery postoperative CT bleeding risk assessment method, an assessment model and an establishment method thereof. The assessment model comprises influence factors points, disease types, sex, hospitalization time, operation sites, operation duration, total influence factor points and a linear score comparison table. The bleeding probability obtained after calculation through the evaluation model provided by the invention is compared with the actual bleeding condition of a patient after an operation in practice, the accuracy and reliability of the evaluation model are verified, and a more accurate theoretical basis is provided for judging whether the patient is subjected to CT examination or not in the early stage after the neurosurgery operation.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of neurosurgery medical technology, and in particular to an evaluation method, an evaluation model and an establishment method thereof for postoperative CT bleeding risk in neurosurgery. Background Art

[0002] Postoperative bleeding is one of the more common complications of elective neurosurgery. Compared with other surgical specialties, neurosurgery generally has longer operation time and a higher risk of complications; most of the operation areas of neurosurgery are located in the brain, and postoperative bleeding is somewhat hidden. When patients show corresponding signs, the amount of bleeding is often large, missing the window of early intervention, and even endangering the patient's life in severe cases. Therefore, in hospitals with conditions, it has become routine to perform CT examinations early after neurosurgery (generally within 6 hours). However, the results of postoperative CT examinations do not meet the expectations of surgeons in most cases. In addition, CT examinations not only increase the patient's radiation exposure and increase the risk of transportation during the examination, but also increase the financial and human costs of patients and medical institutions. Therefore, if we can explore the high-risk factors for postoperative bleeding and accurately judge the necessity of postoperative CT review, we can take into account the safety of patients and reduce radiation exposure and costs. Summary of the invention

[0003] The purpose of the present invention is to establish a model for predicting the probability of postoperative CT bleeding in patients, which provides an accurate basis for determining whether patients undergoing elective neurosurgery should undergo CT examination in the early postoperative period.

[0004] To achieve the above objectives, in a first aspect, the present invention provides a method for establishing an assessment model for the risk of postoperative CT bleeding in neurosurgery, comprising the following steps:

[0005] a. Organize the original data of all patients before, during and after neurosurgery, and classify and count the original data;

[0006] b. Use the univariate logistics regression method in R language to analyze each data type in the original data and screen out the data types of dependent variables with statistical significance. The formula is as follows:

[0007] Logit(pi)=1 / (1+exp(-pi))

[0008] ln(pi / (1-pi))=Beta_0+Beta_1*X_1+…+B_k*K_k, where pi refers to the score of the i-th data type, and Beta_i refers to the percentage value of the i-th data type.

[0009] c. Use multi-factor logistics regression method to analyze the data of all the selected dependent variable data types, and calculate the influencing factors of each dependent variable data type. The formula is as follows:

[0010] Logit(pi)=1 / (1+exp(-pi))

[0011] ln(pi / (1-pi))=Beta_0+Beta_1*X_1+…+B_k*K_k, where pi refers to the score of the i-th dependent variable data type, and Beta_i refers to the percentage value of the i-th dependent variable data type.

[0012] d. Create a nomogram based on the influencing factors of each dependent variable data type to form an evaluation model.

[0013] As a further improvement of the present invention, the original data before, during and after the operation in step a include the patient's disease type, gender, hospitalization time, surgical site, surgical duration, surgical method, lesion side, age, and degree of resection.

[0014] As a further improvement of the present invention, the classification statistics of the original data include statistical calculation of the proportion of the number of people in each type of data among the patient's disease type, gender, hospitalization time, surgical site, surgical duration, surgical method, lesion side, age, and resection degree.

[0015] As a further improvement of the present invention, the method of using a univariate logistics regression method in R language to perform data analysis on each data type in the original data includes using a univariate logistics regression method in R language to perform data analysis on the statistical calculation results of the proportion of each type of data in the patient's disease type, gender, length of hospital stay, surgical site, duration of surgery, surgical method, lesion side, age, and degree of resection, and obtaining a relative risk OR value using a univariate logistics regression method.

[0016] As a further improvement of the present invention, the screening out of the dependent variable data types with statistical significance includes screening out the corresponding data types with relative risk OR values ​​in the range of 0<1 or >1 and defining them as the dependent variable data types with statistical significance.

[0017] As a further improvement of the present invention, the dependent variable data with statistical significance are the patient's disease type, gender, hospitalization time, surgical site, and surgical duration.

[0018] As a further improvement of the present invention, the calculation of the influencing factors of each dependent variable data type in step c includes calculating the relative risk OR value of each dependent variable data type according to the following formula to obtain the score pi of each dependent variable data type,

[0019] The dependent variable data type score pi=(Beta_i / Max_Beta)*100, Beta_i refers to the percentage value of the i-th dependent variable data type, and Max_Beta refers to the maximum percentage value of all dependent variable data types.

[0020] As a further improvement of the present invention, in step d, a nomogram is made according to the influencing factors of each dependent variable data type, and an evaluation model is formed, including making linear straight lines for the values ​​of each dependent variable data type and making a linear score comparison table according to the score pi of each dependent variable data type to form an evaluation model.

[0021] The present invention also provides an evaluation model for the risk of postoperative CT bleeding in neurosurgery, wherein the evaluation model includes influencing factor points, disease type, gender, hospitalization time, surgical site, surgical duration, total influencing factor points, and a linear score comparison table.

[0022] The present invention also provides a method for assessing the risk of postoperative CT bleeding in neurosurgery, comprising the following steps:

[0023] a. Obtain the dependent variable data of the patients before, during and after neurosurgery, including disease type, gender, length of hospital stay, surgical site and duration of surgery;

[0024] b. Compare the specific types of each dependent variable data of the patient with the assessment model of postoperative CT bleeding risk of neurosurgery to obtain the influencing factor points corresponding to each dependent variable data of the patient;

[0025] c. Add up the influencing factor points corresponding to each dependent variable data, and then map them with the linear score comparison table of CT bleeding risk after neurosurgery to obtain the probability of brain hemorrhage in patients after neurosurgery.

[0026] Compared with the prior art, the present invention has the following beneficial effects:

[0027] The present invention provides an assessment model for the risk of postoperative CT bleeding in neurosurgery. The model is based on analysis of CT examination results within 6 hours after surgery and about 24 hours after surgery for 8,840 patients in a hospital in Shanghai. It is found that factors such as disease type, gender, length of hospital stay, surgical site, and duration of surgery have statistically significant effects on postoperative bleeding. The above variables are incorporated into logistics regression analysis, a nomogram is made, and an assessment model is formed. The assessment model is then used to randomly select patients for reverse verification. The verification data shows that the bleeding probability calculated by the assessment model provided by the present invention is compared with the actual situation of postoperative bleeding in actual patients, which verifies the accuracy and reliability of the assessment model of the present invention, and provides a more accurate theoretical basis for determining whether a CT examination should be performed in the early stage after a patient undergoes neurosurgery. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] Figure 1 A schematic diagram of statistical variables and statistical results included in the present invention;

[0029] Figure 2 A flow chart of a method for establishing an assessment model for the risk of postoperative CT bleeding in neurosurgery according to the present invention;

[0030] Figure 3 It is a schematic diagram of the evaluation model of the risk of postoperative CT bleeding in neurosurgery according to the present invention;

[0031] Figure 4 This is a method for assessing the risk of CT bleeding after neurosurgery surgery. DETAILED DESCRIPTION

[0032] The present invention is described in detail below in conjunction with the various embodiments shown in the accompanying drawings, but it should be noted that these embodiments are not limitations of the present invention, and any equivalent transformations or substitutions in functions, methods, or structures made by ordinary technicians in the field based on these embodiments are all within the scope of protection of the present invention.

[0033] This embodiment provides a method for establishing an assessment model for the risk of postoperative CT bleeding in neurosurgery, comprising the following steps:

[0034] a. Organize the original data of all patients before, during and after neurosurgery, and classify and count the original data;

[0035] b. Use the univariate logistics regression method in R language to analyze each data type in the original data and screen out the data types of dependent variables with statistical significance. The formula is as follows:

[0036] Logit(pi)=1 / (1+exp(-pi))

[0037] ln(pi / (1-pi))=Beta_0+Beta_1*X_1+…+B_k*K_k, where pi refers to the score of the i-th data type, and Beta_i refers to the percentage value of the i-th data type.

[0038] c. Use multi-factor logistics regression method to analyze the data of all the selected dependent variable data types, and calculate the influencing factors of each dependent variable data type. The formula is as follows:

[0039] Logit(pi)=1 / (1+exp(-pi))

[0040] ln(pi / (1-pi))=Beta_0+Beta_1*X_1+…+B_k*K_k, where pi refers to the score of the i-th dependent variable data type, and Beta_i refers to the percentage value of the i-th dependent variable data type.

[0041] d. Create a nomogram based on the influencing factors of each dependent variable data type to form an evaluation model.

[0042] like Figure 1 As shown in the figure, based on the collected CT examination results of 8,840 patients within 6 hours after surgery and within 24 hours after surgery, all the patients' disease types, surgical methods, gender, hospitalization time, lesion side, location, duration of surgery, age, degree of resection, etc. were classified. Specifically, the disease types were further subdivided into tumors, functional diseases, vascular diseases and other categories. Among them, tumors were further subdivided into more than ten specific tumor types, and so on, each other type was subdivided into each specific subcategory.

[0043] The raw data before, during and after the operation in step a include the patient's disease type, gender, hospital stay, surgical site, duration of operation, surgical method, lesion side, age, and degree of resection. The classification statistics of the raw data include statistical calculation of the proportion of the number of people in each type of data among the patient's disease type, gender, hospital stay, surgical site, duration of operation, surgical method, lesion side, age, and degree of resection.

[0044] As a further improvement of the present invention, the method of using a univariate logistics regression method in R language to perform data analysis on each data type in the original data includes using a univariate logistics regression method in R language to perform data analysis on the statistical calculation results of the proportion of each type of data in the patient's disease type, gender, length of hospital stay, surgical site, duration of surgery, surgical method, lesion side, age, and degree of resection, and obtaining a relative risk OR value using a univariate logistics regression method.

[0045] The screening of the dependent variable data types with statistical significance includes screening the corresponding data types with relative risk OR values ​​in the range of 0<1 or>1 and defining them as the dependent variable data types with statistical significance. The dependent variable data with statistical significance are the patient's disease type, gender, hospitalization time, surgical site, and surgical duration.

[0046] The calculation of the influencing factors of each dependent variable data type in step c includes calculating the relative risk OR value of each dependent variable data type according to the following formula to obtain the score pi of each dependent variable data type,

[0047] The dependent variable data type score pi=(Beta_i / Max_Beta)*100, Beta_i refers to the percentage value of the i-th dependent variable data type, and Max_Beta refers to the maximum percentage value of all dependent variable data types.

[0048] In the present invention, pi is the score of each subdivided dependent variable data type, which can be obtained through linear regression calculation. In the present invention, the dependent variable data with statistical significance are the patient's disease type, gender, hospitalization time, surgical site, and operation duration.

[0049] In the step d, a nomogram is made according to the influencing factors of each dependent variable data type, and the evaluation model is formed, including making linear lines for the values ​​of each dependent variable data type and making a linear score comparison table according to the scores pi of each dependent variable data type to form the evaluation model.

[0050] The present invention provides an assessment model for the risk of postoperative CT bleeding in neurosurgery, including influencing factor points, disease type, gender, hospital stay, surgical site, surgical duration, total influencing factor points, and a linear score comparison table. In the disease type, the points of functional disease (microvascular decompression) are nearly 0, the points of vascular disease are 24, the points of tumor are 25, and the points of functional disease-epilepsy are 27; in gender, the points of female are nearly 0, and the points of male are 5; in the diseased site, the points of spinal canal are nearly 0, the points of infratentorial are 82.5, the points of sellar region are 85, the points of skull base are 88, and the points of ventricle are nearly 100.

[0051] The present invention also provides a method for assessing the risk of postoperative CT bleeding in neurosurgery, comprising the following steps:

[0052] a. Obtain the dependent variable data of the patients before, during and after neurosurgery, including disease type, gender, length of hospital stay, surgical site and duration of surgery;

[0053] b. Compare the specific types of each dependent variable data of the patient with the assessment model of postoperative CT bleeding risk of neurosurgery to obtain the influencing factor points corresponding to each dependent variable data of the patient;

[0054] c. Add up the influencing factor points corresponding to each dependent variable data, and then map them with the linear score comparison table of CT bleeding risk after neurosurgery to obtain the probability of brain hemorrhage in patients after neurosurgery.

[0055] In order to verify the reliability of the assessment model for the risk of postoperative CT bleeding in neurosurgery provided by the present invention, a portion of the medical records of 8840 patients were randomly selected for verification.

[0056] The specific verification medical records are as follows:

[0057] A male patient with convex meningioma was hospitalized for 15 days and underwent meningioma resection surgery, which took 300 minutes. Then: the patient's disease type score is 25 points, gender score is 5 points, hospitalization time score is 10 points, surgical site score is 88 points, and surgical duration score is 5 points. The total score is 133 points. According to the prediction value at the bottom of the model corresponding figure, the percentage of bleeding in postoperative CT is close to 0.

[0058] Finally, we refer to the CT scan results of the patient after surgery, which showed no bleeding, from which we can know the accuracy of the model's prediction of the probability of surgical CT bleeding.

[0059] On this basis, the remaining more than 100 patients were verified in turn. From the verification results, it can be known that the proportion of patients with a predicted probability of postoperative bleeding less than 10% reached 60%, and the proportion of patients with a predicted probability of postoperative bleeding greater than 80% reached 40%, and the proportion of people in between tended to 0. Further comparison showed that for patients with a bleeding probability of less than 10%, the postoperative CT results showed no bleeding or even if bleeding, it was a very small amount of bleeding, and would not cause brain damage to the patient. For patients with a bleeding probability of more than 80%, after comparing the CT results, more than 97% of the patients had varying degrees of bleeding after surgery. It can be known from this that the accuracy and reliability of the assessment model for postoperative CT bleeding risk in neurosurgery provided by the present invention.

[0060] In addition, as the number of patient medical records increases and big data collection improves, the model calculation method provided by the present invention can further optimize and improve the model, making the score more accurate and the prediction results further improved.

[0061] The series of detailed descriptions listed above are only specific descriptions of feasible implementation methods of the present invention. They are not intended to limit the scope of protection of the present invention. Any equivalent implementation methods or changes that do not deviate from the technical spirit of the present invention should be included in the scope of protection of the present invention.

[0062] It will be apparent to those skilled in the art that the invention is not limited to the details of the exemplary embodiments described above and that the invention can be implemented in other specific forms without departing from the spirit or essential features of the invention. Therefore, the embodiments should be considered exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description, and it is intended that all variations falling within the meaning and scope of the equivalent elements of the claims be included in the invention. Any reference numeral in a claim should not be considered as limiting the claim to which it relates.

[0063] In addition, it should be understood that although the present specification is described according to implementation modes, not every implementation mode contains only one independent technical solution. This description of the specification is only for the sake of clarity. Those skilled in the art should regard the specification as a whole. The technical solutions in each embodiment may also be appropriately combined to form other implementation modes that can be understood by those skilled in the art.

Claims

1. A method for establishing a model for assessing the risk of postoperative CT bleeding in neurosurgery. Features: The following steps are included: a. Organize the original data of all patients before, during and after neurosurgery, and classify and count the original data; b. Use the univariate logistics regression method in R language to analyze each data type in the original data and screen out the data types of dependent variables with statistical significance. The formula is as follows: Logit(pi)=1 / (1+exp(-pi)) ln(pi / (1-pi))=Beta_0+Beta_1*X_1+…+B_k*K_k, where pi refers to the score of the i-th data type, and Beta_i refers to the percentage value of the i-th data type. c. Use multi-factor logistics regression method to analyze the data of all the selected dependent variable data types, and calculate the influencing factors of each dependent variable data type. The formula is as follows: Logit(pi)=1 / (1+exp(-pi)) ln(pi / (1-pi))=Beta_0+Beta_1*X_1+…+B_k*K_k, where pi refers to the score of the i-th dependent variable data type, and Beta_i refers to the percentage value of the i-th dependent variable data type. d. Create a nomogram based on the influencing factors of each dependent variable data type to form an evaluation model.

2. The method for establishing an assessment model for the risk of postoperative CT bleeding in neurosurgery according to claim 1, Features: The original data before, during and after the operation in step a include the patient's disease type, gender, length of hospital stay, surgical site, duration of operation, surgical method, lesion side, age and extent of resection.

3. The method for establishing an assessment model for the risk of postoperative CT bleeding in neurosurgery according to claim 2, Features: The classification statistics for the original data include statistical calculation of the proportion of the number of people in each type of data in terms of patient disease type, gender, length of hospital stay, surgical site, duration of surgery, surgical method, lesion side, age, and degree of resection.

4. The method for establishing an assessment model for the risk of postoperative CT bleeding in neurosurgery according to claim 3, Features: The method of using a univariate logistics regression method in R language to perform data analysis on each data type in the original data includes using a univariate logistics regression method in R language to perform data analysis on the statistical calculation results of the proportion of each type of data in the patient's disease type, gender, length of hospital stay, surgical site, duration of surgery, surgical method, lesion side, age, and degree of resection, and obtaining the relative risk OR value using the univariate logistics regression method.

5. The method for establishing an assessment model for the risk of postoperative CT bleeding in neurosurgery according to claim 4, Features: The method of screening out the dependent variable data type with statistical significance includes screening out the corresponding data type with relative risk OR value in the range of 0<1 or >1 and defining it as the dependent variable data type with statistical significance.

6. The method for establishing an assessment model for the risk of postoperative CT bleeding in neurosurgery according to claim 5, Features: The statistically significant dependent variable data are the patient's disease type, gender, length of hospital stay, surgical site, and duration of surgery.

7. The method for establishing an assessment model for the risk of postoperative CT bleeding in neurosurgery according to claim 6, Features: The calculation of the influencing factors of each dependent variable data type in step c includes calculating the relative risk OR value of each dependent variable data type according to the following formula to obtain the score pi of each dependent variable data type, The dependent variable data type score pi=(Beta_i / Max_Beta)*100, Beta_i refers to the percentage value of the i-th dependent variable data type, and Max_Beta refers to the maximum percentage value of all dependent variable data types.

8. The method for establishing an assessment model for the risk of postoperative CT bleeding in neurosurgery according to claim 7, Features: In the step d, a nomogram is made according to the influencing factors of each dependent variable data type, and the evaluation model is formed, including making linear lines for the values ​​of each dependent variable data type and making a linear score comparison table according to the scores pi of each dependent variable data type to form the evaluation model.

9. A model for assessing the risk of postoperative CT bleeding in neurosurgery. Features: The evaluation model includes influencing factor points, disease type, gender, hospitalization time, surgical site, surgical duration, total influencing factor points, and a linear score comparison table.

10. A method for assessing the risk of postoperative CT bleeding in neurosurgery. Features: The following steps are included: a. Obtain the dependent variable data of the patients before, during and after neurosurgery, including disease type, gender, length of hospital stay, surgical site and duration of surgery; b. Compare the specific types of each dependent variable data of the patient with the assessment model of postoperative CT bleeding risk of neurosurgery to obtain the influencing factor points corresponding to each dependent variable data of the patient; c. Add up the influencing factor points corresponding to each dependent variable data, and then map them with the linear score comparison table of CT bleeding risk after neurosurgery to obtain the probability of brain hemorrhage in patients after neurosurgery.