Prediction model construction method and application of primary corpus callosum degeneration related pneumonia

Abstract

The invention discloses a construction method and application of a prediction model for primary corpus callosum-related pneumonia, and the construction method comprises the following steps: obtaining clinical data and laboratory test results of patients with primary corpus callosum degeneration; carrying out single-factor analysis and screening related factors having statistical significance with the primary corpus callosum degeneration-related pneumonia; the method comprises the following steps: carrying out dimensionality reduction treatment on related factors with statistical significance of primary corpus callosum degeneration-related pneumonia, and carrying out multi-factor logistic regression analysis on the related factors subjected to dimensionality reduction treatment to determine independent related factors of primary corpus callosum degeneration-related pneumonia; and according to the independent related factors of the primary corpus callosum degeneration-related pneumonia, constructing a prediction model of the primary corpus callosum degeneration-related pneumonia. The prediction model constructed by the invention has good prediction capability, and clinicians can identify high-risk patients with primary corpus callosum degeneration related pneumonia as soon as possible and perform intervention and treatment as soon as possible.

Description

Technical Field

[0001] The present invention relates to the technical field of medical information prediction models, and in particular to a prediction model construction method and application of primary corpus callosum degeneration-related pneumonia. Background Art

[0002] Primary degeneration of the corpus callosum, also known as Marchiafava-Bignami disease (MBD), is a critical neuropsychiatric emergency primarily associated with chronic alcoholism. Its pathological features are symmetrical demyelination, necrosis, cystic degeneration, and atrophy of the central layer of the corpus callosum. Most patients with primary degeneration of the corpus callosum have a poor prognosis, resulting in disability or death, which severely affects their quality of life, endangers their physical health, shortens their lifespan, and increases the economic burden on their families and society.

[0003] Previous studies have shown that pneumonia is one of the most common complications in patients with primary degeneration of the corpus callosum and an independent risk factor for poor prognosis in patients with primary degeneration of the corpus callosum. If clinicians wait until patients meet the diagnostic criteria for pneumonia before intervening, the adverse effects of pneumonia on the prognosis of patients with primary degeneration of the corpus callosum have already occurred. Therefore, early identification of patients at high risk for primary degeneration of the corpus callosum-related pneumonia and provision of preventive interventions can help reduce the incidence of primary degeneration of the corpus callosum-related pneumonia.

[0004] However, there is currently a lack of prediction tools for primary corpus callosum degeneration-associated pneumonia, and there is an urgent need for a simple, convenient and fast assessment tool to predict primary corpus callosum degeneration-associated pneumonia. Summary of the Invention

[0005] In response to the shortcomings of existing technical methods, the purpose of the present invention is to comprehensively analyze the clinical manifestations and laboratory test results of patients with primary corpus callosum degeneration with and without pneumonia, screen the related factors of primary corpus callosum degeneration-related pneumonia using elastic network regression analysis, determine the independent related factors of primary corpus callosum degeneration-related pneumonia through multivariate logistic regression, and construct a prediction model for primary corpus callosum degeneration-related pneumonia based on a nomogram. The present invention provides clinicians with a simple, convenient, and rapid prediction and assessment tool for primary corpus callosum degeneration-related pneumonia, helping clinicians to identify patients at high risk of primary corpus callosum degeneration-related pneumonia at an early stage, intervene as soon as possible, and improve patient prognosis.

[0006] In order to achieve the above object, the present invention provides a method for constructing a prediction model for primary corpus callosum degeneration-related pneumonia, comprising:

[0007] Clinical data and laboratory test results of patients with primary corpus callosum degeneration were obtained, and univariate analysis was performed to screen for statistically significant factors associated with primary corpus callosum degeneration-related pneumonia;

[0008] The statistically significant related factors of primary corpus callosum degeneration-related pneumonia were processed for dimensionality reduction, and the related factors after dimensionality reduction were subjected to multivariate logistic regression analysis to determine the independent related factors of primary corpus callosum degeneration-related pneumonia;

[0009] A prediction model for primary corpus callosum degeneration-related pneumonia was constructed based on the independent related factors of primary corpus callosum degeneration-related pneumonia.

[0010] Furthermore, the dimensionality reduction processing of the statistically significant related factors of primary corpus callosum degeneration-related pneumonia is performed, and the multivariate logistic regression analysis of the related factors after dimensionality reduction processing is performed includes:

[0011] Whether patients with primary corpus callosum degeneration had pneumonia was used as the dependent variable. Elastic network regression analysis and LASSO regression analysis were used to screen the related factors of primary corpus callosum degeneration-related pneumonia after dimensionality reduction, and the union of the two factors was taken.

[0012] The combined data were directly entered into the multivariate logistic regression analysis.

[0013] Furthermore, the patients with primary corpus callosum degeneration must meet the inclusion criteria and not the exclusion criteria;

[0014] Inclusion criteria: age ≥ 18 years, regardless of gender; first hospitalization, or only the first hospitalization for those with multiple hospitalizations; head CT and / or MRI examinations confirming the presence of corpus callosum demyelination, necrosis, atrophy, and other lesions;

[0015] Exclusion criteria: patients with cirrhosis and autoimmune diseases.

[0016] Furthermore, the clinical data include gender, age, clinical manifestations, time from onset to admission, Glasgow Coma Scale score at admission, and concurrent and concurrent diseases;

[0017] Laboratory test results included white blood cell count, total red blood cell count, hemoglobin, total platelet count, sodium, potassium, chloride, random blood glucose, alanine aminotransferase, aspartate aminotransferase, and gamma-glutamyltransferase.

[0018] Furthermore, based on the independent related factors of primary corpus callosum degeneration-associated pneumonia, a prediction model for primary corpus callosum degeneration-associated pneumonia was constructed, including:

[0019] The independent related factors of primary corpus callosum degeneration-related pneumonia were statistically analyzed, and a nomogram prediction model for primary corpus callosum degeneration-related pneumonia was constructed.

[0020] Furthermore, after constructing the prediction model for primary corpus callosum degeneration-associated pneumonia, its predictive ability was evaluated:

[0021] The predictive ability of the prediction model was evaluated by the area under the receiver operating characteristic curve, the consistency of the prediction model was evaluated by drawing a calibration curve, and the clinical application efficacy of the prediction model was evaluated using decision curve analysis.

[0022] The present invention also provides a prediction model for primary corpus callosum degeneration-related pneumonia, which is obtained using the above-mentioned construction method.

[0023] The present invention also provides the clinical application of the above-mentioned prediction model for primary corpus callosum degeneration-related pneumonia.

[0024] The present invention also provides a prediction system for primary corpus callosum degeneration-related pneumonia, comprising a memory, a processor, and a program stored in the memory and executable on the processor. When the memory executes the program, the prediction model is executed to implement the following steps:

[0025] Independently associated factors for primary corpus callosum degeneration-associated pneumonia in collected and / or infused patients;

[0026] The independent related factors of the patient's primary corpus callosum degeneration-related pneumonia are brought into the prediction model for calculation to obtain the prediction results;

[0027] Output a conclusion on whether the patient is at risk for primary corpus callosum degeneration-associated pneumonia.

[0028] The present invention also provides a computer-readable storage medium storing the above-mentioned program.

[0029] Compared with the prior art, the present invention has the following beneficial effects:

[0030] The prediction model constructed by the present invention has good predictive ability, which can enable clinicians to identify patients at high risk of primary corpus callosum degeneration-related pneumonia as early as possible, intervene and treat as soon as possible, efficiently allocate medical resources, reduce the incidence of complications and sequelae of the disease, improve patients' clinical prognosis, and reduce family and socioeconomic burdens. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0032] Figure 1 shows a plot of the optimal alpha value for elastic net regression analysis;

[0033] Figure 2 shows the coefficient trajectory plot for the elastic net regression analysis;

[0034] Figure 3 Cross-validation curves for elastic net regression analysis are shown;

[0035] Figure 4 A common nomogram prediction model for primary corpus callosum degeneration-associated pneumonia is shown;

[0036] Figure 5 A network dynamic nomogram prediction model for primary corpus callosum degeneration-associated pneumonia is shown;

[0037] Figure 6 The ROC curve of the nomogram prediction model for primary corpus callosum degeneration-associated pneumonia is shown;

[0038] Figure 7 The calibration curve of the nomogram prediction model for primary corpus callosum degeneration-associated pneumonia is shown;

[0039] Figure 8 The DCA curve of the nomogram prediction model for primary corpus callosum degeneration-related pneumonia is shown;

[0040] Figure 9 The ROC curve of the internal validation group of the prediction model for primary corpus callosum degeneration-associated pneumonia is shown;

[0041] Figure 10 The ROC curve of the external validation group of the prediction model for primary corpus callosum degeneration-associated pneumonia is shown;

[0042] Figure 11 A prediction example of the network dynamic nomogram prediction model for primary corpus callosum degeneration-associated pneumonia is shown. DETAILED DESCRIPTION

[0043] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the specific embodiments of the present invention and the accompanying drawings. Obviously, the embodiments described are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0044] A method for constructing a prediction model for primary corpus callosum degeneration-associated pneumonia comprises the following steps:

[0045] Step 1: Data collection:

[0046] (1) Patients with primary degeneration of the corpus callosum admitted to the Department of Neurology, Affiliated Hospital of Zunyi Medical University from 2012 to 2025 were selected as research subjects. Patients who were treated from 2013 to 2021 were used as the derivation group, and patients who were treated from 2012 and 2022 to 2025 were used as the internal validation group. In addition, "MarchiafavaBignami" was used as the keyword to search PubMed and Web of Science, and published case reports of primary degeneration of the corpus callosum were used as the external validation group. The patients were divided into pneumonia group and non-pneumonia group according to whether they had pneumonia.

[0047] (2) The collected clinical data included gender, age, time from onset to admission, Glasgow Coma Scale (GCS) score at admission, clinical manifestations (presence of impaired consciousness, disordered thinking, delusions, hallucinations, apathy, mutism, aphasia, blurred vision, vomiting, sensory impairment, hypertonia, increased tendon reflexes, positive pathological signs, first epileptic seizure, status epilepticus, multiple epileptic seizures, urinary and fecal incontinence, unsteady walking, dizziness, headache, poor diet), and concurrent diseases;

[0048] (3) The collected laboratory test results included white blood cell count (WBC), total red blood cell count, hemoglobin, total platelet count, sodium, potassium, chloride, random blood glucose, alanine aminotransferase (ALT), aspartate aminotransferase (AST), and gamma-glutamyltransferase (GGT).

[0049] (4) Diagnostic criteria for primary degeneration of the corpus callosum: 1) History of long-term heavy drinking, or meeting the diagnostic criteria for alcohol dependence in ICD-10 / DSM IV, or history of malnutrition; 2) Acute, subacute, or chronic onset, with clinical symptoms and signs of neuropsychiatric abnormalities; 3) Imaging: cranial CT shows symmetrical low-density lesions in the corpus callosum, or cranial MRI shows symmetrical high signals in the corpus callosum on T2WI and FLAIR, with or without extracallosal white matter lesions; 4) Exclusion of other causes of neuropsychiatric diseases.

[0050] (5) Diagnostic criteria for pneumonia associated with primary degeneration of the corpus callosum: Patients with primary degeneration of the corpus callosum present with clinical manifestations and auxiliary examination results that meet the diagnostic criteria for pneumonia.

[0051] (6) Inclusion criteria: 1) Age ≥ 18 years, regardless of gender; 2) First hospitalization, or only the first hospitalization for those with multiple hospitalizations; 3) Patients with primary corpus callosum degeneration underwent head CT and / or MRI examinations, confirming the presence of corpus callosum demyelination, necrosis, atrophy and other lesions.

[0052] (7) Exclusion criteria: 1) Patients with concurrent liver cirrhosis; 2) Patients with concurrent autoimmune diseases such as systemic lupus erythematosus.

[0053] Step 2: Statistical analysis of data:

[0054] (1) Normality tests were performed on measurement data. Measurement data that conformed to normal distribution were expressed as mean ± standard deviation, and the differences between the two groups were compared using independent sample t-tests. Measurement data that were not normally distributed were expressed as median and 25% and 75% interquartile ranges, and the differences between the two groups were compared using nonparametric tests (Mann-Whitney U test). Enumeration data were expressed as frequencies and percentages, and the differences between the two groups were compared using chi-square tests.

[0055] In this example, the baseline data for the primary degeneration of the corpus callosum group with pneumonia and the primary degeneration of the corpus callosum group without pneumonia are as follows. Compared with the group without pneumonia, the primary degeneration of the corpus callosum patients in the pneumonia group had fewer symptoms of confusion and headache, and more symptoms of mutism, status epilepticus, urinary and fecal incontinence, poor diet, impaired consciousness, increased muscle tone, lower Glasgow Coma Scale (GCS) scores, and higher white blood cell counts and blood glucose levels. See Table 1 for details.

[0056] Table 1 Comparison of clinical characteristics of patients with primary corpus callosum degeneration between the pneumonia group and the non-pneumonia group within the derivation group

[0057]

[0058]

[0059] GCS, Glasgow Coma Scale; IQR, interquartile range;

[0060] Fisher's exact probability method was used to obtain P values.

[0061] *P<0.05

[0062] (2) The factors significantly associated with primary corpus callosum degeneration-related pneumonia (P<0.05) were selected from the univariate analysis, including confusion, mutism, status epilepticus, urinary and fecal incontinence, headache, poor diet, impaired consciousness, increased muscle tone, GCS score, white blood cell count, and blood glucose level. These factors were then entered into elastic network regression analysis and LASSO regression analysis to screen for factors associated with primary corpus callosum degeneration-related pneumonia.

[0063] like Figure 1-3 As shown in the figure, elastic network regression analysis found the optimal hyperparameters through cross-validation. The optimal alpha value corresponding to the best performance index was 0.85, log(lambda min) value was 0.008, and log(lambda 1se) value was 0.055. The related factors of primary corpus callosum degeneration-related pneumonia were determined to be GCS score, confusion, status epilepticus, urinary and fecal incontinence, headache, white blood cell count, and blood glucose level.

[0064] LASSO regression analysis revealed that the related factors for primary corpus callosum degeneration-related pneumonia were GCS score, confusion, status epilepticus, urinary and fecal incontinence, headache and white blood cell count.

[0065] The relevant factors screened out by elastic network regression analysis and LASSO regression analysis were combined, namely, GCS score, confusion, status epilepticus, urinary and fecal incontinence, headache, white blood cell count, and blood glucose level. Multivariate logistic regression analysis was performed using the direct entry method. The independent related factors of pneumonia in patients with primary corpus callosum degeneration were determined based on the odds ratio (OR) and 95% confidence interval (CI), as shown in Table 2.

[0066] Table 2 Independent factors associated with pneumonia associated with primary corpus callosum degeneration

[0067]

[0068] GCS, Glasgow Coma Scale; OR, odds ratio; CI, confidence interval;

[0069] *P<0.05

[0070] Finally, GCS score, confusion and white blood cell count were identified as independent related factors for primary corpus callosum degeneration-associated pneumonia.

[0071] Step 3: Build a prediction model:

[0072] According to the independent related factors determined in step 2, SPSS29.0 software was used for statistical analysis of the data, and R software (version 4.4.2) was used to construct the nomogram prediction model. R software packages such as car, rms, pROC, and rmda were used to construct the following Figure 4 The common nomogram prediction model for primary corpus callosum degeneration-associated pneumonia is shown in Figure 1. The R software packages rms and DynNom were used to construct the model. Figure 5 The network dynamic nomogram prediction model for primary corpus callosum degeneration-associated pneumonia is shown. The web link for this prediction model is https: / / mbdresearch.shinyapps.io / MBDPneumonia / .

[0073] Step 4: Evaluation of the predictive ability of the prediction model:

[0074] The area under the receiver operating characteristic (ROC) curve (AUC) was used to evaluate the predictive ability of the nomogram prediction model. The consistency of the prediction model was evaluated by drawing a calibration curve. The clinical application efficacy of the nomogram prediction model was evaluated using decision curve analysis (DCA).

[0075] The ROC curve of the nomogram prediction model for primary corpus callosum degeneration-related pneumonia is shown as follows Figure 6 As shown in the figure, its AUC is 0.875, indicating that the nomogram prediction model has good predictive ability, its Youden index is 0.663, and the optimal cutoff value is 0.502. That is, when the prediction model of primary corpus callosum degeneration-related pneumonia predicts that a patient has a risk of primary corpus callosum degeneration-related pneumonia ≥ 0.502, the patient can be considered to be at high risk of primary corpus callosum degeneration-related pneumonia, with a sensitivity of 76.32% and a specificity of 90.00%.

[0076] Figure 7 The calibration curve of the nomogram prediction model is close to the ideal curve, indicating that the predicted incidence of pneumonia in the prediction model of primary corpus callosum degeneration-related pneumonia is consistent with the actual incidence. From the DCA diagram and data analysis, it can be seen that the nomogram prediction model is far away from the two gray extreme curves, indicating that the nomogram prediction model has good clinical application efficiency. Figure 8 .

[0077] A total of 114 patients with primary corpus callosum degeneration who were treated in 2012 and 2022 to 2025 were selected as the internal validation group. Compared with the derivation group, the patients with primary corpus callosum degeneration in the internal validation group had higher GCS scores, lower γ-glutamyltransferase levels, more frequent electrolyte disturbances, and less frequent apathy, status epilepticus, and multiple epileptic seizures. There were no significant differences between the two groups in other items (see Table 3). The AUC of the internal validation group was 0.834, confirming that the nomogram had good prognostic prediction ability (see Table 3). Figure 9 .

[0078] Table 3 Comparison of clinical characteristics of patients with primary corpus callosum degeneration in the derivation group and the internal validation group

[0079]

[0080]

[0081] GCS, Glasgow Coma Scale; IQR, interquartile range; Fisher's exact probability method was used to obtain P values.

[0082] *P<0.05

[0083] We searched Pubmed and Web of Science using the keyword "Marchiafava Bignami," and found 376 articles in Pubmed and 640 articles in Web of Science. We removed duplicate articles, irrelevant articles, reviews, articles not related to alcohol poisoning, and abstracts. We also included references on primary corpus callosum degeneration in the retrieved articles. Finally, we included 80 articles and 83 MBD cases as the external validation group. The AUC of this prediction model in the external validation group was 0.950, which once again confirmed that this nomogram has good prognostic prediction ability. Figure 10 .

[0084] According to the contents described in steps 3 and 4, select the published case report. A patient with primary corpus callosum degeneration has a GCS score of 6, confusion, and a normal white blood cell count. Select the value of each item on the dynamic nomogram for the prediction of primary corpus callosum degeneration-related pneumonia and click Predict. In the prediction probability graph on the right, read the predicted risk of primary corpus callosum degeneration-related pneumonia for this patient as 0.860, which is greater than 0.502. This patient can be considered at high risk of primary corpus callosum degeneration-related pneumonia. See Figure 11. The journal article "COVID-19-induced acute loss of consciousness in Marchiafava-Bignamidisease: a case report" published in the Journal of International Medical Research, 2024, Volume 52, Issue 4, Page 1, shows that the patient's final diagnosis was primary corpus callosum degeneration-associated pneumonia.

[0085] In summary, the present invention constructs a prediction model for primary corpus callosum degeneration-associated pneumonia, and uses multidimensional model evaluation methods such as the area under the curve of the receiver operating characteristic curve, calibration curve, decision curve analysis, internal validation, and external validation, all of which show that the prediction model of the present invention has good predictive ability for predicting primary corpus callosum degeneration-associated pneumonia. At the same time, a common nomogram that does not rely on electronic devices and network connections and a simple, convenient, and fast network dynamic nomogram are established as evaluation tools, which help clinicians identify high-risk patients for primary corpus callosum degeneration-associated pneumonia as early as possible and intervene and treat them as soon as possible.

[0086] Finally, it should be noted that the above is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art can still modify the technical solutions described in the aforementioned embodiments or make equivalent substitutions for some of the technical features therein. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for constructing a prediction model for primary corpus callosum degeneration-associated pneumonia, characterized in that: include, Clinical data and laboratory test results of patients with primary corpus callosum degeneration were obtained, and univariate analysis was performed to screen for statistically significant factors associated with primary corpus callosum degeneration-related pneumonia; The statistically significant related factors of primary corpus callosum degeneration-related pneumonia were processed for dimensionality reduction, and the related factors after dimensionality reduction were subjected to multivariate logistic regression analysis to determine the independent related factors of primary corpus callosum degeneration-related pneumonia; A prediction model for primary corpus callosum degeneration-related pneumonia was constructed based on the independent related factors of primary corpus callosum degeneration-related pneumonia.

2. The method for constructing a prediction model for primary corpus callosum degeneration-associated pneumonia according to claim 1, wherein: The method of performing dimensionality reduction processing on the statistically significant related factors of primary corpus callosum degeneration-related pneumonia and performing multivariate logistic regression analysis on the related factors after dimensionality reduction processing includes: Whether patients with primary corpus callosum degeneration had pneumonia was used as the dependent variable. Elastic network regression analysis and LASSO regression analysis were used to screen the related factors of primary corpus callosum degeneration-related pneumonia after dimensionality reduction, and the union of the two factors was taken. The combined data were directly entered into the multivariate logistic regression analysis.

3. The method for constructing a prediction model for primary corpus callosum degeneration-associated pneumonia according to claim 1, wherein: The patients with primary corpus callosum degeneration must meet the inclusion criteria and not the exclusion criteria; Inclusion criteria: age ≥ 18 years, regardless of gender; first hospitalization, or only the first hospitalization for those with multiple hospitalizations; head CT and / or MRI examinations confirming the presence of corpus callosum demyelination, necrosis, atrophy, and other lesions; Exclusion criteria: patients with cirrhosis and autoimmune diseases.

4. The method for constructing a prediction model for primary corpus callosum degeneration-associated pneumonia according to claim 1, wherein: The clinical data included gender, age, clinical manifestations, time from onset to admission, Glasgow Coma Scale score at admission, and concurrent and concurrent diseases; Laboratory test results included white blood cell count, total red blood cell count, hemoglobin, total platelet count, sodium, potassium, chloride, random blood glucose, alanine aminotransferase, aspartate aminotransferase, and gamma-glutamyltransferase.

5. The method for constructing a prediction model for primary corpus callosum degeneration-associated pneumonia according to claim 1, wherein: According to the independent related factors of primary corpus callosum degeneration-related pneumonia, the prediction model of primary corpus callosum degeneration-related pneumonia was constructed, including: The independent related factors of primary corpus callosum degeneration-related pneumonia were statistically analyzed, and a nomogram prediction model for primary corpus callosum degeneration-related pneumonia was constructed.

6. The method for constructing a prediction model for primary corpus callosum degeneration-associated pneumonia according to claim 1, wherein: After constructing the prediction model for primary corpus callosum degeneration-associated pneumonia, its predictive ability was also evaluated: The predictive ability of the prediction model was evaluated by the area under the receiver operating characteristic curve, the consistency of the prediction model was evaluated by drawing a calibration curve, and the clinical application efficacy of the prediction model was evaluated using decision curve analysis.

7. A prediction model for primary corpus callosum degeneration-associated pneumonia, characterized in that: The method is obtained by the construction method according to any one of claims 1 to 6.

8. A clinical application of the prediction model for pneumonia associated with primary corpus callosum degeneration according to claim 7.

9. A system for predicting pneumonia associated with primary corpus callosum degeneration, comprising a memory, a processor, and a program stored in the memory and executable on the processor, characterized in that: When the memory runs the program, the prediction model according to claim 8 is executed to implement the following steps: Independently associated factors for primary corpus callosum degeneration-associated pneumonia in collected and / or infused patients; The independent related factors of the patient's primary corpus callosum degeneration-related pneumonia are brought into the prediction model for calculation to obtain the prediction results; Output a conclusion on whether the patient is at risk for primary corpus callosum degeneration-associated pneumonia. 10 . A computer-readable storage medium storing the program according to claim 9 .

Images