Brain tumor postoperative infection evaluation model construction method and system
By constructing a Fisher discriminant model based on multiple parameters of cerebrospinal fluid, the problems of long diagnosis time and low accuracy in intracranial infection diagnosis in existing technologies are solved, efficient intracranial infection diagnosis and prognosis assessment are achieved, and the targeted treatment is improved.
Patent Information
- Application Number
- CN202510712164.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-05-29
- Publication Date
- 2025-10-03
AI Technical Summary
Existing diagnostic methods for intracranial infection are time-consuming and have limited sensitivity. Clinical judgments that rely on a single or a few parameters are irrational, leading to delayed treatment and poor prognosis.
A Fisher discriminant model based on multiple parameters of cerebrospinal fluid was constructed, integrating CSF parameters such as total protein, glucose, chloride, white blood cell count, and multinuclear cell count to generate a discriminant function for early diagnosis and prediction of intracranial infection.
The diagnostic accuracy and treatment guidance of intracranial infection were improved. The diagnostic sensitivity of the model was 88.68% and the specificity was 77.01%. Cerebrospinal fluid parameters were significantly correlated with the infection control time, and the prognosis assessment was effective.
Smart Images

Figure CN120748704A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of medical detection technology, and in particular to a postoperative infection assessment model for brain tumor surgery based on cerebrospinal fluid (CSF) test indicators, which is used for rapid diagnosis of intracranial infection and prediction of patient prognosis. Background Art
[0002] Postoperative intracranial infection (ICI) is a significant complication associated with poor prognosis in patients with brain tumors, primarily caused by pathogens invading the brain parenchyma and meninges. Although its incidence has significantly decreased with advances in aseptic techniques and the use of prophylactic antibiotics, complete elimination of postoperative intracranial infection remains a challenge. A retrospective analysis of patients undergoing craniotomy for brain tumors showed an overall incidence of 6.8%. Therefore, timely and accurate identification of intracranial infection is crucial for initiating targeted treatment and improving patient outcomes.
[0003] Currently, the standard for diagnosing intracranial infection is cerebrospinal fluid (CSF) culture, a valuable tool for guiding antimicrobial susceptibility testing and molecular epidemiological studies. However, its detection requires at least one day and has limited sensitivity, with a false-negative rate of 10%-40%, primarily due to the reduced culture-positive rate caused by the use of prophylactic antibiotics during the perioperative period. Furthermore, other methods, such as polymerase chain reaction (PCR) and metagenomic next-generation sequencing (mNGS), are increasingly being used in clinical laboratories to achieve independent diagnosis without relying on culture. However, the positive detection rates of these methods still rely on the presence of pathogens, similar to CSF culture. Relying solely on CSF etiology results to diagnose intracranial infection may lead to delayed treatment, ultimately adversely affecting patient prognosis.
[0004] Based on this, clinical practice often relies on clinical features (such as the classic triad of bacterial meningitis—fever, headache, and nuchal stiffness) and laboratory parameters to initiate empirical anti-infective therapy, but such approaches have limitations in both sensitivity and specificity. Trauma, bleeding, and postoperative sterile inflammation not only cause changes in routine blood tests and procalcitonin (PCT) levels but also interfere with CSF cytology and biochemistry, complicating the diagnosis of intracranial infection. Notably, CSF white blood cell counts, protein levels, and glucose (CSF / serum) ratios can remain within normal ranges in some patients with shunt-related meningitis, suggesting the complexity of intracranial infection. Relying solely on a single or limited set of parameters combined with subjective clinical judgment may not be rational. Therefore, a rapid, objective, multi-parameter model is urgently needed to improve diagnostic accuracy and guide treatment. Summary of the Invention
[0005] To address the irrationality of existing technologies that rely solely on a single or limited number of parameters combined with subjective clinical judgment, this paper provides a method for constructing a postoperative infection assessment model for brain tumors. This method, based on the Fisher discriminant model integrating multiple cerebrospinal fluid parameters, is used for the early diagnosis and prognosis of postoperative infection after brain tumor surgery. By analyzing data from 389 patients, nine significantly different CSF parameters were identified, and the model was constructed and its clinical value validated.
[0006] The technical solutions of the present invention are as follows:
[0007] In a first aspect, the present invention provides a method for constructing a postoperative infection assessment model for brain tumors, comprising the following steps:
[0008] (1) Fisher discriminant analysis was performed to integrate the results of cerebrospinal fluid total protein (C-TP), glucose (C-Glu), chloride (C-Cl-), leukocyte count (C-leukocytes), multinucleated cell count (C-multinucleated), multinucleated cell ratio (C-multinucleated%), color, turbidity, and Pandy test to generate a discriminant function;
[0009] (2) Prognosis prediction: Analyze the correlation between the levels of cerebrospinal fluid parameters and the time of infection control.
[0010] According to a specific embodiment of the present invention, the discriminant function of step (1) is:
[0011] F = -74.225 + 1.248 x C-TP + 0.743 x C-Glu + 1.11 x C-Cl - + 0.002 x C-leukocytes - 0.002 x C-multinucleated + 0.069 x C-multinucleated % - 2.282 x color + 3.662 x turbidity + 6.684 x Pandy test.
[0012] According to a specific embodiment of the present invention, step (1) further includes: diagnostic verification, evaluating consistency with clinical diagnosis;
[0013] According to a specific embodiment of the present invention, the diagnostic verification method is verified by ROC curve. The area under the curve AUC for diagnosing intracranial infection is 0.927 for white blood cell count and 0.909 for multinuclear cell count; among which: the AUC values of cerebrospinal fluid total protein, Cl-, glucose, glucose ratio and multinuclear cell ratio are 0.846, 0.645, 0.799, 0.752 and 0.747, respectively.
[0014] According to a specific embodiment of the present invention, in step (2), prognosis prediction using cerebrospinal fluid total protein, white blood cell count, increased ratio of multinuclear cells and decreased glucose level are significantly correlated with prolonged infection control time.
[0015] In a second aspect, the present invention provides a model system for evaluating infection after brain tumor surgery, comprising cerebrospinal fluid biochemical analysis reagents, cell counting reagents, and the above-constructed evaluation model, for outputting infection risk levels and prognosis prediction results in real time.
[0016] According to a specific embodiment of the present invention, the cerebrospinal fluid biochemical analysis reagents include: a cerebrospinal fluid / urine total protein (CSF / UP) assay kit (pyrophenol red method) (Kinsel), a glucose detection kit (hexokinase method) (Meikang), an AU biochemical analysis system special reagent electrolyte internal standard solution (BECKMAN COULTER), and a buffer ISE BUFFER (BECKMANCOULTER).
[0017] According to a specific embodiment of the present invention, the cell counting reagent includes: a diluent for blood cell analysis CELLPACK DFL, a hemolytic agent for blood cell analysis SULFOLYSER, and a hemolytic agent for blood cell analysis Lysercell WDF.
[0018] According to a specific embodiment of the present invention, the infection risk levels include high, medium, and low.
[0019] According to a specific embodiment of the present invention, the infection level is determined by calculating the number of "+"s for each patient based on five significant indicators; the five significant indicators include: cerebrospinal fluid total protein (C-TP), cerebrospinal fluid glucose (C-Glu), cerebrospinal fluid leukocyte count (C-leukocytes), multinucleated cell count (C-multinucleated), and multinucleated cell percentage (C-multinucleated%);
[0020] 4 to 5 "+"s indicate a high-risk group;
[0021] 2 to 3 "+"s indicate the medium-risk group;
[0022] ≤1 "+" indicates low-risk group.
[0023] Secondly, the application of the above-mentioned brain tumor postoperative infection assessment model system in the evaluation of brain tumor postoperative infection and prognosis prediction is provided.
[0024] This patent is funded by the Taishan Scholar Young Experts Project (No.tsqn202312366); the Joint Innovation Fund of Shandong Natural Science Foundation (ZR2023LZL011); and the Shandong Provincial Traditional Chinese Medicine Science and Technology Project (M2023-013).
[0025] Beneficial effects of the present invention
[0026] 1. There were statistically significant differences in cerebrospinal fluid test parameter levels between the intracranial infection group and the non-infection group.
[0027] The present invention proposes that the total protein, white blood cell count, polymorphonuclear cell count, and polymorphonuclear cell ratio in the cerebrospinal fluid of intracranial infection patients are significantly increased (all P < 0.0001), while the levels of chloride ion (Cl-), glucose, and cerebrospinal fluid / serum glucose ratio (C-Glu / S-Glu) are significantly decreased (P values are 0.0015, <0.0001, 0.0016 respectively). In addition, the serum procalcitonin (PCT) and C-reactive protein (CRP) in intracranial infection patients are significantly increased (all P < 0.0001). At the same time, there are also significant differences in cerebrospinal fluid color, turbidity, and Pandy test results between the two groups (all P < 0.0001). ROC curve analysis shows that the diagnostic efficacy of cerebrospinal fluid white blood cell count is the best (AUC = 0.927), followed by polymorphonuclear cell count (AUC = 0.909).
[0028] 2. Construction of an evaluation model for postoperative brain tumor infection based on cerebrospinal fluid test indicators
[0029] Abnormal cerebrospinal fluid parameters were screened out based on variable comparison analysis (Mann-Whitney U test or chi-square test), and finally 9 variables were included to construct a Fisher discriminant model. The established discriminant function for intracranial infection is:
[0030] <F = -74.225 + 1.248 × total protein + 0.743 × glucose + 1.11 × Cl- + 0.002 × white blood cell count - 0.002 × polymorphonuclear cell count + 0.069 × polymorphonuclear cell ratio - 2.282 × color + 3.662 × turbidity + 6.684 × Pandy test>.
[0031] 3. The constructed model has good consistency with clinical practice.
[0032] The sensitivity of the model for suspected infection cases is 88.68%, and the specificity is 77.01%. Kappa analysis shows that the discriminant model has good consistency with clinical diagnosis results (κ = 0.625, P < 0.0001).
[0033] 4. Cerebrospinal fluid parameters are significantly correlated with the time of infection control.
[0034] Kaplan-Meier survival analysis shows that an increase in total protein, white blood cell count, polymorphonuclear cell ratio in cerebrospinal fluid and a decrease in glucose level are significantly correlated with an extended time of infection control (P < 0.0001). The infection control time in the low-risk group is significantly shorter than that in the medium- and high-risk groups (P < 0.0001), indicating that these parameters are closely related to the prognosis of intracranial infection. Description of the Drawings
[0035] Figure 1 Research phase flow chart and classification scheme;
[0036] Figure 2 Distribution of pathogens in cerebrospinal fluid of patients with intracranial infection; A represents the composition of bacteria isolated from cerebrospinal fluid culture specimens sent for examination after surgery; B represents the distribution of multidrug-resistant bacteria;
[0037] Figure 3 Comparison of cerebrospinal fluid (CSF)-related parameter levels between the infection group and the non-infection group; A includes C-TP (CSF total protein), C-Cl- (CSF chloride ion), C-Glu (CSF glucose), and C-Glu / S-Glu (CSF / serum glucose ratio); B includes C-leukocytes (CSF white blood cell count), C-multinucleated (multinucleated cell count), and C-multinucleated% (multinucleated cell percentage); C includes PCT (procalcitonin) and CRP (C-reactive protein);
[0038] Figure 4 ROC curves were used to analyze the diagnostic efficacy of CSF parameters for postoperative intracranial infection in patients with brain tumors. A shows the ROC curves for the following CSF parameters for diagnosing postoperative intracranial infection: C-TP (total protein), C-Cl- (chloride ion), C-Glu (glucose), and C-Glu / S-Glu (cerebrospinal fluid / serum glucose ratio); B shows the ROC curves for CSF leukocyte count (C-leukocytes), multinucleated cell count (C-multinucleated), and multinucleated cell percentage (C-multinucleated%); C shows the AUC (area under the curve), 95% confidence interval (95% CI), and P value for each parameter.
[0039] Figure 5 Flowchart of follow-up management of patients with intracranial infection;
[0040] Figure 6 The KM survival curve shows the correlation between cerebrospinal fluid parameters and the prognosis of intracranial infection; A is the grouping based on CSF parameter indicators: including cerebrospinal fluid total protein (C-TP), cerebrospinal fluid glucose (C-Glu), cerebrospinal fluid leukocyte count (C-leukocytes), multinucleated cell count (C-multinucleated), and multinucleated cell percentage (C-multinucleated%); B is the grouping based on risk stratification: high-risk, intermediate-risk, and low-risk groups. DETAILED DESCRIPTION
[0041] The following examples are used to illustrate the present invention but are not intended to limit the scope of the present invention. Unless otherwise specified, the technical means used in the examples are conventional means known to those skilled in the art, and the raw materials used are commercially available products.
[0042] Example 1: Diagnostic and prognostic value of combined cerebrospinal fluid parameters for intracranial infection in patients with brain tumors after neurosurgery
[0043] 1. Study population and inclusion criteria
[0044] A total of 694 patients who underwent surgery for brain tumors at Shandong Cancer Hospital, affiliated with Shandong First Medical University, between February 2020 and April 2024 were enrolled. Inclusion criteria included patients aged ≥16 years, diagnosed with an intracranial tumor, and undergoing neurosurgery with cerebrospinal fluid culture. Exclusion criteria included: age <16 years; no surgical treatment; use of >1 dose of antibiotics; absence of a brain tumor; incomplete data; and comorbidities with other infectious, immune, psychiatric, or metabolic diseases. Based on the inclusion and exclusion criteria, 389 eligible patients underwent cerebrospinal fluid culture. Of the 92 patients with positive cultures, 35 were excluded due to surgical wound infection, scalp infection, actinomycete or fungal infection, unclear diagnosis, or specimen contamination. Ultimately, 57 patients with positive cerebrospinal fluid cultures were defined as the "infection group." Among the 297 patients with negative cerebrospinal fluid culture, 157 patients with normal blood routine indicators and no clinical symptoms were defined as the "non-infection group"; the remaining 140 patients were defined as the "suspected group" because they met at least one of the following criteria: clinical manifestations of intracranial infection (body temperature >38°C, headache, meningeal irritation signs or focal neurological symptoms); abnormal blood routine and / or inflammatory mediators (procalcitonin PCT, C-reactive protein CRP). Finally, 389 cases that met the criteria were screened (57 cases in the infection group, 157 cases in the non-infection group, and 140 cases in the suspected group). The flow chart of the research subjects is shown below. Figure 1 Table 1 summarizes the clinical and laboratory characteristics, including age, sex, smoking and drinking status, hypertension, diabetes, and tumor type.
[0045] The diagnostic criteria for intracranial infection were determined when the patient met any of the following diagnostic criteria: (1) clinical diagnosis: high fever, intracranial hypertension, turbid or purulent cerebrospinal fluid, white blood cell count ≥10×106, blood glucose <2.2mmol / L, cerebrospinal fluid glucose / blood glucose ≤0.4; (2) positive results of clinical diagnostic specimen smears, drainage tube tips, implants, and cerebrospinal fluid bacterial culture. This study fully respected the 1964 Declaration of Helsinki and was reviewed and approved by the Ethics Committee of Shandong Cancer Hospital Affiliated to Shandong First Medical University and Shandong Academy of Medical Sciences. Written informed consent was obtained from all participants or their legal representatives. All information obtained from patient files was kept confidential, and no patient identification information was captured.
[0046] Table 1 Clinical characteristics of the total cohort
[0047]
[0048]
[0049] 2. Detection methods and data collection
[0050] A postoperative infection assessment model system for brain tumors includes cerebrospinal fluid biochemical analysis reagents, cell counting reagents, and the above-constructed assessment model, which is used to output infection risk levels and prognosis prediction results in real time.
[0051] Reagents for cerebrospinal fluid biochemical analysis include: cerebrospinal fluid / urine total protein (CSF / UP) assay kit (pyrophenol red method) (Kinsel), glucose assay kit (hexokinase method) (Meikang), AU biochemical analysis system-specific reagent electrolyte internal standard solution (BECKMAN COULTER), and buffer ISE BUFFER (BECKMAN COULTER).
[0052] Cell counting reagents include: CELLPACK DFL diluent for blood cell analysis, SULFOLYSER hemolytic agent for blood cell analysis, and Lysercell WDF hemolytic agent for blood cell analysis.
[0053] CSF cultures included direct plate culture and automated broth enrichment. Fresh samples were inoculated onto Columbia blood agar and chocolate agar plates and cultured directly under 5% CO2 aerobic conditions. Microbial growth observed within 24–48 h was identified using a BRUKER microflex MALDI-TOF / TOF mass spectrometer (BRUKER, MA, USA). Simultaneously, the remaining CSF was transferred to culture bottles for enrichment over a 5-day period. Positive signals were screened for false positives using positivity time threshold and growth curve slope analysis. Positive broths were Gram-stained, subcultured onto agar plates, and incubated for 24 h; colonies were confirmed using the same MALDI-TOF platform. Bacterial drug susceptibility analysis was performed using a BD Phoenix M50 automated microbiology system (Dickinson and Company, Becton, NJ, USA). All procedures were performed according to the manufacturer's instructions and the Clinical Laboratory Standards Institute BC standards.
[0054] For cerebrospinal fluid (CSF) physicochemical parameters, total protein, glucose level, and chloride ion were measured using a Beckman Coulter analyzer AU5800 (Beckman Coulter, Brea, CA, USA); cell count and differential were determined using a Sysmex XN-550 (Sysmex, Kobe, Japan). Qualitative tests such as CSF colorimetry, turbidity, and Pandy results were performed by senior laboratory technicians. Interpretation standards were based on the operating procedures of clinical examinations in my country. A negative result of a qualitative test (colorless, transparent, and (−) Pandy test) was designated as "0," and a positive result of a qualitative test (colored, opaque, and (+) Pandy test) was designated as "1."
[0055] For hematological parameters, serum procalcitonin (PCT) was quantified using a Roche Cobas e801 analyzer (Roche Diagnostics GmbH, Mannheim, Germany); levels below 0.05 ng / mL were considered normal. High-sensitivity C-reactive protein (CRP) and blood cell count and differential were analyzed using a Mindray BC-7500 fully automated hematology analyzer (Mindray Medical, Shenzhen, China). The manufacturer's reference value for CRP was <10 mg / L.
[0056] 3. Mathematical statistical analysis
[0057] Statistical analyses were performed using SPSS 26.0 statistical software (SPSS Inc., Chicago, IL, USA) and GraphPad Prism version 9.5 (GraphPad Software Inc., San Diego, CA, USA). For comparisons of nonnormal continuous variables, the Mann-Whitney U test was used for independent sample comparisons between two groups. Categorical data were compared using the chi-square test or Fisher's exact test. Diagnostic performance was assessed using the area under the receiver operating characteristic (ROC) curve (AUC). A discriminant model was established using Fisher discriminant analysis, and suspected cases were included in the model to verify the validity of the discriminant function. The Kappa test was used to assess the consistency of the model's predictions with the clinically confirmed results. Infection control time was defined as the interval from the diagnosis of intracranial infection to clinical resolution of infection or discharge. Kaplan-Meier survival curves were plotted, and the relationship between infection control time and the level of different parameters was analyzed using the log-rank test. Data are expressed as percentages or medians (interquartile range, IQR). P < 0.05 was considered statistically significant.
[0058] 4. Microbiological characteristics of infected samples
[0059] In the 57 patients in the infection group, a total of 68 pathogens were identified through cerebrospinal fluid culture. Among them, 8 patients had multiple pathogen infections (5 were infected with two pathogens and 3 were infected with three pathogens), and the rest were infected with a single pathogen. Among the 42 cases of Gram-positive bacterial infection, coagulase-negative Staphylococcus predominated (25 cases, accounting for 36.76% of the total positive cultures), followed by viridans streptococci (13 cases, accounting for 19.12%). Among the 26 Gram-negative bacteria, Acinetobacter and Enterobacter were common pathogens (8 cases each, accounting for 11.76% of the total positive cultures). For details on the composition of isolated bacterial species and the distribution of multidrug-resistant bacteria in cerebrospinal fluid specimens, please see Figure 2 .
[0060] 5. Comparison of differences in cerebrospinal fluid test parameters between the infection group and the non-infection group
[0061] In the infection group, the total protein, white blood cell count, polymorphonuclear cell count and polymorphonuclear cell ratio in cerebrospinal fluid were significantly increased (all P < 0.0001), while the levels of chloride ion (Cl-), glucose and cerebrospinal fluid / serum glucose ratio (C-Glu / S-Glu) were significantly decreased (P values were 0.0015, < 0.0001 and 0.0016, respectively) ( Figure 3 A, 3B). In addition, serum procalcitonin (PCT) and C-reactive protein (CRP) were significantly elevated in patients with intracranial infection (both P < 0.0001; Figure 3 Table 2 shows that there were significant differences in CSF color, turbidity, and Pandy test results between the two groups (all P < 0.0001).
[0062] Table 2 Comparison of categorical data (color, turbidity and Pandy test) between the infected and non-infected groups
[0063]
[0064] 6. Diagnostic efficacy of cerebrospinal fluid parameters for intracranial infection in patients with brain tumors after surgery
[0065] The present invention evaluated the diagnostic value of the above parameters for intracranial infection using ROC curves. The results showed that the cerebrospinal fluid white blood cell count had the best diagnostic efficacy (AUC = 0.927), followed by the multinuclear cell count (AUC = 0.909). The AUC values for cerebrospinal fluid total protein, Cl-, glucose, glucose ratio (C-Glu / S-Glu), and multinuclear cell ratio were 0.846, 0.645, 0.799, 0.752, and 0.747, respectively. Figure 4 .
[0066] 7. Construction of discriminant model and diagnostic verification
[0067] Abnormal cerebrospinal fluid parameters screened based on variable comparison analysis (Mann-Whitney U test or chi-square test), and finally 9 variables were included to construct a Fisher discriminant model. The established discriminant function for intracranial infection is: <F = -74.225 + 1.248×total protein + 0.743×glucose + 1.11×Cl- + 0.002×white blood cell count - 0.002×polymorphonuclear cell count + 0.069×proportion of polymorphonuclear cells - 2.282×color + 3.662×turbidity + 6.684×Pandy test>.
[0068] To verify the effectiveness of the model, 140 cases in the "suspected group" screened were substituted into this discriminant model, and the calculated results were compared with the final clinical diagnosis (53 cases of infection and 87 cases of non-infection). The results showed that the diagnostic sensitivity of this model was 88.68% and the specificity was 77.01%. Kappa analysis indicated that there was a good consistency between the discriminant model and the clinical diagnosis results (κ = 0.625, P < 0.0001; Table 3)
[0069] Table 3 Sensitivity and specificity of Fisher discriminant function
[0070]
[0071] 8. Association between cerebrospinal fluid parameter indicators and prognosis of intracranial infection
[0072] 124 patients with intracranial infection confirmed by cerebrospinal fluid culture and Fisher discriminant analysis were included in the follow-up analysis of the time to infection control. Among them, 2 cases were excluded due to missing data, see Figure 5 . Finally, 122 infected patients were divided into two groups according to the median level of each cerebrospinal fluid index for Kaplan-Meier survival analysis. The time to infection control in the groups with elevated total protein, white blood cell count, polymorphonuclear cell count and proportion of polymorphonuclear cells in cerebrospinal fluid, and the group with decreased glucose was significantly prolonged (P values were 0.0009, <0.0001, <0.0001, 0.0037, 0.0001 respectively), suggesting that these parameters are closely related to the prognosis of intracranial infection.
[0073] The cerebrospinal fluid parameters that lead to prolonged infection control time are marked as "+" to represent high risk, and the number of "+" for each patient is calculated based on 5 significant indicators. The indicators mentioned are: cerebrospinal fluid total protein (C-TP), cerebrospinal fluid glucose (C-Glu), cerebrospinal fluid leukocyte count (C-leukocytes), multinucleated cell count (C-multinucleated) and multinucleated cell percentage (C-multinucleated%); according to the number of "+", the patients are divided into: high-risk group (4 to 5 "+"), medium-risk group (2 to 3 "+") and low-risk group (≤1 "+"). The infection control time of the low-risk group was significantly shorter than that of the medium and high-risk groups (P<0.0001). See Figure 6 .
Claims
1. A method for constructing a postoperative infection assessment model for brain tumors, characterized in that: The following steps are involved: (1) Fisher discriminant analysis was performed to integrate the results of cerebrospinal fluid total protein (C-TP), glucose (C-Glu), chloride (C-Cl-), leukocyte count (C-leukocytes), multinucleated cell count (C-multinucleated), multinucleated cell ratio (C-multinucleated%), color, turbidity, and Pandy test to generate a discriminant function; (2) Prognosis prediction: Analyze the correlation between the levels of cerebrospinal fluid parameters and the time of infection control.
2. The evaluation model construction method according to claim 1, characterized in that: The discriminant function of step (1) is: F = -74.225 + 1.248 x C-TP + 0.743 x C-Glu + 1.11 x C-Cl - + 0.002 x C-leukocytes - 0.002 x C-multinucleated + 0.069 x C-multinucleated % - 2.282 x color + 3.662 x turbidity + 6.684 x Pandy test.
3. The evaluation model construction method according to claim 1, characterized in that: Step (1) also includes: diagnostic verification, evaluating consistency with clinical diagnosis; Preferably, the diagnostic verification method is verified by ROC curve verification, and the area under the curve AUC for diagnosing intracranial infection is 0.927 for white blood cell count and 0.909 for multinuclear cell count; among which: the AUC values of cerebrospinal fluid total protein, Cl-, glucose, glucose ratio and multinuclear cell ratio are 0.846, 0.645, 0.799, 0.752 and 0.747, respectively.
4. The method for constructing an evaluation model according to any one of claims 1 to 3, characterized in that: Step (2) prognosis prediction using cerebrospinal fluid total protein, white blood cell count, increased proportion of multinuclear cells and decreased glucose level were significantly associated with prolonged infection control time.
5. A model system for evaluating infection after brain tumor surgery, characterized in that: The device comprises a cerebrospinal fluid biochemical analysis reagent, a cell counting reagent and an evaluation model constructed according to any one of claims 1 to 4, and is used for outputting infection risk levels and prognosis prediction results in real time.
6. The brain tumor postoperative infection assessment model system according to claim 5, characterized in that: Cerebrospinal fluid biochemical analysis reagents include: cerebrospinal fluid / urine total protein determination kit, glucose detection kit, AU biochemical analysis system special reagent electrolyte internal standard solution, buffer ISE BUFFER.
7. The brain tumor postoperative infection assessment model system according to claim 5, characterized in that: Cell counting reagents include: CELLPACK DFL diluent for blood cell analysis, SULFOLYSER hemolytic agent for blood cell analysis, and Lysercell WDF hemolytic agent for blood cell analysis.
8. The brain tumor postoperative infection assessment model system according to any one of claims 5 to 7, characterized in that: The infection risk levels shown include high, intermediate, and low.
9. The brain tumor postoperative infection assessment model system according to any one of claims 5 to 7, characterized in that: Determination of infection grade: The number of "+" marks for each patient was calculated based on five significant indicators: cerebrospinal fluid total protein (C-TP), cerebrospinal fluid glucose (C-Glu), cerebrospinal fluid leukocyte count (C-leukocytes), multinucleated cell count (C-multinucleated), and multinucleated cell percentage (C-multinucleated%). 4 to 5 "+"s indicate a high-risk group; 2 to 3 "+"s indicate the medium-risk group; ≤1 "+" indicates low-risk group.
10. Use of the brain tumor postoperative infection assessment model system according to any one of claims 5 to 9 in the assessment of infection and prognosis prediction after brain tumor surgery.
