Non-small cell lung cancer brain metastasis prediction method, system and device

Abstract

The invention provides a method, a system and equipment for predicting brain metastasis of non-small cell lung cancer (NSCLC), which are specially designed for brain metastasis (BM) risk of non-small cell lung cancer (NSCLC) patients, and are used for constructing a model for predicting the brain metastasis risk of the non-small cell lung cancer patients by adopting various machine learning algorithms based on lung cancer patient data of an SEER database. The performance of the gradient elevator in a test set and a verification set is superior to that of other classification models, the model also obtains the highest accuracy, sensitivity and Kappa value and has the lowest Brier score, and the method can also provide a visual tool which can be used for clinically helping doctors make decisions, early identifying high-risk patients and planning treatment and nursing in advance.

Description

Technical Field

[0001] The present invention relates to the technical field of cancer brain metastasis diagnosis, and in particular to a method, system and device for predicting brain metastasis of non-small cell lung cancer. Background Art

[0002] Brain metastasis (BM) is one of the most serious neurological complications of cancer, contributing significantly to morbidity and mortality in affected patients. Common primary malignancies associated with BM include lung cancer, breast cancer, and melanoma. In non-small cell lung cancer (NSCLC), approximately 10%–25% of patients present with BM at first presentation, and approximately 50% will develop BM during treatment. The incidence of BM in patients with NSCLC at first presentation has been reported to be 10% to 20%, and can be as high as 40% during the clinical course. The prognosis for patients with BM is generally poor, with a median survival time of less than 1 year, and patients often experience poor quality of life and prognosis.

[0003] With advances in treatment, many patients with non-small cell lung cancer and brain metastases now benefit from stereotactic radiosurgery, a key component of multimodal treatment aimed at improving treatment outcomes. However, some brain metastases from non-small cell lung cancer may be latent and asymptomatic. In addition, the prognosis of patients with such brain metastases is generally poor. Therefore, timely detection of brain metastases is necessary for subsequent examinations and treatment planning. Therefore, identifying risk factors for brain metastases is crucial to improve predictive evaluation and guide clinical treatment.

[0004] In the field of cutting-edge artificial intelligence methods, machine learning (ML), including logistic regression (LR), classification and regression trees (CART), random forest (RF), support vector machine (SVM), K nearest neighbor (KNN), gradient boosting machine (G-Brain Metastasis), and extreme gradient boosting (XGBOOST) algorithms, has become an innovative and widely used tool. Its importance lies in its ability to enhance the predictive accuracy of diagnostic processes and prognostic assessments. ML algorithms are able to autonomously learn knowledge from training data sets, promote dynamic interactions between variables, and identify possible key predictor variables. By identifying patterns within data sets, these algorithms establish an optimized relationship between outcomes and potential predictor variables. This autonomous improvement in decision-making ability significantly improves the effectiveness of predictions, showing a significant improvement over other methods. Summary of the invention

[0005] The purpose of the present invention is to provide a method, system and device for predicting brain metastasis of non-small cell lung cancer, develop a prediction model through a machine learning algorithm, evaluate the risk probability of brain metastasis of non-small cell lung cancer, and establish a convenient visualization tool to help clinicians predict the risk of brain metastasis of non-small cell lung cancer.

[0006] In order to achieve the above object, the present invention provides a method for predicting brain metastasis in patients with non-small cell lung cancer, comprising:

[0007] Data acquisition, acquiring patient data based on the SEER database, wherein the patient data includes a set of variables of patient baseline information;

[0008] Data preprocessing, excluding variables with missing values, and using the LASSO algorithm to remove patient data with insignificant variables;

[0009] Model construction: randomly divide the screened patient data into training sets, validation sets, and test sets, and build multiple machine learning models to achieve the best prediction performance;

[0010] Statistical analysis: training machine learning models based on training sets and validation sets, using test sets to evaluate the above training results based on ROC curve area, calibration curve, and model performance parameter evaluation model, and selecting the model with the best performance.

[0011] Furthermore, the baseline information of the patient data includes age, gender, marital status, histological type, tumor size, tumor side, surgical history, radiotherapy history, chemotherapy history, T stage, N stage, distant lymph node metastasis, brain metastasis, and lung metastasis.

[0012] Furthermore, according to the screening of the LASSO algorithm, the baseline information of the patient data includes: age, marital status, histological type, tumor size, surgical history, radiotherapy history, chemotherapy history, T stage, N stage, distant lymph node metastasis, brain metastasis, and lung metastasis.

[0013] Furthermore, the machine learning models include logistic regression models, classification and regression trees, random forests, support vector machines, K nearest neighbors, gradient boosting machines, and extreme gradient boosting machines.

[0014] Furthermore, based on statistical analysis, the AUC values ​​of the gradient boosting machine model in the validation set and test set were 0.8276 and 0.8301 respectively; the model performance parameters were accuracy: 0.8752, sensitivity: 0.890, specificity: 0.5595, Kappa: 0.224; accuracy: 0.8768, sensitivity: 0.8991, specificity: 0.5826, Kappa: 0.2280, respectively. The gradient boosting machine model was the model with the best performance.

[0015] Furthermore, a visualization tool was established based on the existing model, and the patient's clinical information was input to predict the patient's risk of brain metastasis.

[0016] On the other hand, the present invention also provides a system for predicting brain metastasis of patients with non-small cell lung cancer, comprising:

[0017] A data acquisition module, which acquires patient data based on the SEER database, wherein the patient data includes a set of variables of patient baseline information;

[0018] Data preprocessing module, excluding variables with missing values ​​and using LASSO algorithm to remove patient data with insignificant variables;

[0019] The model building module randomly divides the screened patient data into training sets, validation sets, and test sets, and builds multiple machine learning models to achieve the best prediction performance;

[0020] The statistical analysis module trains the machine learning model based on the training set and the validation set, uses the test set and evaluates the above training results based on the ROC curve area, calibration curve, and model performance parameter evaluation model to select the model with the best performance.

[0021] On the other hand, the present invention further provides a storage device, wherein the storage medium stores a plurality of instructions, and the instructions are suitable for loading by a processor to execute the steps in the above-mentioned method for predicting brain metastasis in patients with non-small cell lung cancer.

[0022] The present invention provides a method, system and device for predicting brain metastasis of non-small cell lung cancer, which are specially designed for the risk of brain metastasis (BM) of patients with non-small cell lung cancer (NSCLC). Based on lung cancer patient data in the SEER database, a model for predicting the risk of brain metastasis of patients with non-small cell lung cancer is constructed by using multiple machine learning algorithms. Tests show that the performance of the gradient boosting machine in the test set and the validation set is better than other classification models. The model also achieves the highest accuracy, sensitivity and Kappa value, and has the lowest Brier score. The present invention can also provide a visualization tool that can be used to help doctors make decisions clinically, identify high-risk patients at an early stage, and plan treatment and care in advance. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0024] Figure 1 The present invention is a flowchart of a method for predicting brain metastasis of non-small cell lung cancer according to an embodiment of the present invention.

[0025] Figure 2 The system framework diagram of a non-small cell lung cancer brain metastasis prediction system according to an embodiment of the present invention.

[0026] Figure 3It is a Lasso regression analysis diagram of an embodiment of the present invention.

[0027] Figure 4 is the ROC curve of each machine learning model of an embodiment of the present invention. DETAILED DESCRIPTION

[0028] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments 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 creative work are within the scope of protection of the present invention.

[0029] In the following description, reference is made to “some embodiments”, which describe a subset of all possible embodiments, but it will be understood that “some embodiments” may be the same subset or different subsets of all possible embodiments and may be combined with each other without conflict.

[0030] If similar descriptions of "first / second" appear in the application documents, the following instructions are added. In the following description, the terms "first\second\third" involved are merely used to distinguish similar objects and do not represent a specific ordering of the objects. It can be understood that "first\second\third" can be interchanged in a specific order or sequence where permitted, so that the embodiments of the present application described herein can be implemented in an order other than that illustrated or described herein.

[0031] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art to which this application belongs. The terms used herein are only for the purpose of describing the embodiments of this application and are not intended to limit this application.

[0032] The poor prognosis of lung cancer with brain metastasis has always attracted much attention. A large amount of literature has explored the prognostic factors of brain metastasis (BM), including age, sex, histological grade, treatment, marital status, distant metastasis, and tumor size. However, most studies focus on the survival assessment of patients with lung cancer and BM. The evaluation of risk factors for BM in patients with non-small cell lung cancer (NSCLC) is crucial to identify high-risk patients, which is necessary for considering candidates for prophylactic cranial irradiation (PCI) or other prophylactic treatments. Previous studies have identified multiple risk factors for BM in NSCLC, including young age, non-squamous cell carcinoma, and disease stage. In addition, EGFR and ALK gene mutations and advanced disease have also been identified as independent factors for the development of BM. High N stage is also considered a risk factor for BM, and lymph node infiltration is associated with an increased risk of BM. Further studies have shown that high serum CEA levels, NSE, and CA125 are also independent risk factors for BM. In patients with small cell lung cancer, higher T stage, higher NLR, earlier chest radiotherapy, and fewer chemotherapy cycles are associated with an increased risk of BM. Consistent with the above findings, our study also demonstrated the important role of tumor size and bone metastasis in BM of NSCLC patients.

[0033] In the prior art, the development of predictive models for distant organ metastasis in lung cancer patients has the potential to identify individuals with clinical features that are more susceptible to metastasis. Currently, many predictive models for BM in patients with non-small cell lung cancer (NSCLC) focus mainly on clinical factors. However, previously reported models have limitations of poor sensitivity and specificity, posing a major challenge to predictive ability. The results of the study suggest that clinicians should consider clinical pathological features when formulating diagnostic and treatment algorithms. The present invention not only helps to identify high-risk patients, but also promotes the formulation of individualized assessments and personalized examination measures. Ultimately, the present invention can extend the life expectancy of patients, improve the quality of life, and reduce the economic burden on society and families.

[0034] The following describes an immune checkpoint inhibitor cardiotoxicity assessment method, system and storage device according to an embodiment of the present invention with reference to the accompanying drawings. First, a immune checkpoint inhibitor cardiotoxicity assessment method according to an embodiment of the present invention will be described with reference to the accompanying drawings.

[0035] Figure 1 FIG. 1 is a flow chart of a method for predicting brain metastasis of non-small cell lung cancer according to an embodiment of the present invention. Figure 1 As shown, the prediction method includes the following steps:

[0036] S1 data acquisition: acquiring patient data based on the SEER database, wherein the patient data includes a set of variables of patient baseline information.

[0037] S2 data preprocessing, excluding variables with missing values, and using the LASSO algorithm to remove patient data with statistically insignificant variables.

[0038] S3 model construction randomly divides the screened patient data into training sets, validation sets, and test sets, and builds multiple machine learning models to achieve the best predictive performance.

[0039] S4 statistical analysis: train the machine learning model based on the training set and the validation set, use the test set and evaluate the above training results based on the ROC curve area, calibration curve, and model performance parameter evaluation model to select the model with the best performance.

[0040] Specifically, in step S1 , the patient data is derived from the Surveillance, Epidemiology, and End Results (SEER) program (www.seer.cancer.gov).

[0041] In a specific embodiment, the inclusion criteria for patient data include: (1) patients diagnosed between January 2010 and December 2020; (2) location and morphology classified as lung and bronchus according to ICD-0-3 / WHO-2008; (3) ICD-0-3 behavior code indicating malignancy; (4) histological type determined as non-small cell lung cancer according to the International Classification of Diseases Oncology (Third Edition) code; (5) the presence of a malignant primary lesion.

[0042] Exclusion criteria for patient data included: (1) lack of specificity for T stage or tumor in situ (Tis) or not detected (T0); (2) lack of specificity for N stage; and (3) lack of specificity for metastasis.

[0043] In one embodiment, the clinical pathological records of NSCLC patients are extracted by SEER*stat software (version 8.4.2). The final extracted patient data baseline information includes age, gender, marital status, histological type, tumor size, tumor side, surgical history, radiotherapy history, chemotherapy history, T stage, N stage, distant lymph node metastasis, brain metastasis, and lung metastasis.

[0044] Specifically, in step S2, variables with missing values ​​in patient data are first excluded, and then the least absolute shrinkage and selection operator (LASSO) regression analysis is used to remove variables that are not statistically significant.

[0045] Figure 3 This is a Lasso regression analysis diagram of an embodiment of the present invention, where the left figure is the evaluation of the variability of the variable coefficients, and the right figure is the process of determining the optimal λ parameter in the Lasso regression model using the cross-validation method

[0046] After screening by the LASSO algorithm, the baseline information of the patient data included: age, marital status, histological type, tumor size, surgical history, radiotherapy history, chemotherapy history, T stage, N stage, distant lymph node metastasis, brain metastasis, and lung metastasis.

[0047] Specifically, in step S3, this embodiment randomly divides 39,930 patients into three parts: a training set (60% of the participants), a validation set (20% of the participants), and a test set (20% of the participants). This embodiment constructs seven machine learning models based on the above data, including logistic regression (LR), classification and regression tree (CART), random forest (RF), support vector machine (SVM), K nearest neighbor (KNN), gradient boosting machine (GBM) and extreme gradient boosting (XGBOOST) to achieve the best prediction performance.

[0048] In step S4, the continuous variables of this example are expressed as medians, and the categorical variables are expressed as numbers and percentages. The accuracy of the training cohort model was tested on the data of the test and validation cohorts. The area under the ROC curve was used to distinguish the model capabilities in the two data sets. All statistical analyses, model construction, and validation in this study were based on R software (version 4.1.3)

[0049] The predictive performance of the model was evaluated by the receiver operating characteristic (ROC) curve, and the higher the area under the curve (AUC) value, the stronger the discrimination ability.

[0050] Figure 4 : It is the ROC curve of each machine learning model of an embodiment of the present invention. The left figure is the ROC curve of the validation cohort, and the right figure is the ROC curve of the test cohort. As shown in the figure, the AUC values ​​in the validation set are: 0.7818, 0.7286, 0.7824, 0.7250, 0.7341, 0.8276 and 0.8227 respectively. In the test set, the AUC values ​​are: 0.7907, 0.7165, 0.7793, 0.7425, 0.7333, 0.8301 and 0.8295 respectively.

[0051] In this embodiment, the model performance parameters of each machine learning model are shown in the following table:

[0052]

[0053]

[0054] Based on the above analysis and statistical analysis, the AUC values ​​of the gradient boosting machine model in the validation set and test set were 0.8276 and 0.8301 respectively; the model performance parameters were accuracy: 0.8752, sensitivity: 0.890, specificity: 0.5595, Kappa: 0.224; accuracy: 0.8768, sensitivity: 0.8991, specificity: 0.5826, Kappa: 0.2280 respectively. The gradient boosting machine model was the model with the best performance.

[0055] In a specific embodiment, in order to facilitate the use of medical staff at work, the present invention uses a mobile device or a computer as an electronic device for running the method of the present invention, and the electronic device includes: a processor and a memory for storing executable instructions of the processor. Among them, the executable instructions stored in the memory are configured to run a visualization program of a non-small cell lung cancer brain metastasis prediction method of the present invention, and the visualization program can communicate data with other terminals, servers or other forms of devices to complete the background program calculation.

[0056] Through the visualization tool, clinicians can input the clinical information of a typical patient into the model, such as a 75-year-old male, single, adenocarcinoma, main bronchus, tumor size 61 mm, T4N3, no surgery or chemotherapy, history of radiotherapy, no lymph node metastasis and lung metastasis. The model predicts that the patient's BM risk is 57.48%, indicating a high BM risk.

[0057] Figure 2 FIG. 1 is a system framework diagram of a non-small cell lung cancer brain metastasis prediction system according to an embodiment of the present invention. Figure 2 As shown, a brain metastasis prediction system for patients with non-small cell lung cancer comprises:

[0058] A data acquisition module 1 is used to acquire patient data based on the SEER database, wherein the patient data includes a set of variables of patient baseline information;

[0059] Data preprocessing module 2, excluding variables with missing values, and using the LASSO algorithm to remove patient data with insignificant variables;

[0060] Model building module 3, randomly divides the screened patient data into training set, validation set and test set, and builds multiple machine learning models to achieve the best prediction performance;

[0061] Statistical analysis module 4 trains the machine learning model based on the training set and the validation set, uses the test set and evaluates the above training results based on the ROC curve area, calibration curve, and model performance parameter evaluation model to select the model with the best performance.

[0062] Although the embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and cannot be understood as limiting the present invention. A person of ordinary skill in the art can change, modify, replace and modify the above embodiments within the scope of the present invention. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for predicting brain metastasis in patients with non-small cell lung cancer, characterized in that: include: Data acquisition, acquiring patient data based on the SEER database, wherein the patient data includes a set of variables of patient baseline information; Data preprocessing, excluding variables with missing values, and using the LASSO algorithm to remove patient data with insignificant variables; Model construction: randomly divide the screened patient data into training sets, validation sets, and test sets, and build multiple machine learning models to achieve the best prediction performance; Statistical analysis: training machine learning models based on training sets and validation sets, using test sets to evaluate the above training results based on ROC curve area, calibration curve, and model performance parameter evaluation model, and selecting the model with the best performance.

2. The method for predicting brain metastasis in patients with non-small cell lung cancer according to claim 1, characterized in that: The baseline information of the patient data includes age, gender, marital status, histological type, tumor size, tumor side, surgical history, radiotherapy history, chemotherapy history, T stage, N stage, distant lymph node metastasis, brain metastasis, and lung metastasis.

3. The method for predicting brain metastasis of patients with non-small cell lung cancer according to claim 2, characterized in that: According to the screening of the LASSO algorithm, the baseline information of the patient data includes: age, marital status, histological type, tumor size, surgical history, radiotherapy history, chemotherapy history, T stage, N stage, distant lymph node metastasis, brain metastasis, and lung metastasis.

4. The method for predicting brain metastasis of patients with non-small cell lung cancer according to claim 3, characterized in that: The machine learning models include logistic regression models, classification and regression trees, random forests, support vector machines, K nearest neighbors, gradient boosting machines, and extreme gradient boosting machines.

5. The method for predicting brain metastasis of patients with non-small cell lung cancer according to claim 4, characterized in that: Based on statistical analysis, the AUC values ​​of the gradient boosting model in the validation set and test set are 0.8276 and 0.8301 respectively; the model performance parameters are accuracy: 0.8752, sensitivity: 0.890, specificity: 0.5595, Kappa: 0.224; Accuracy: 0.8768, Sensitivity: 0.8991, specificity: 0.5826, Kappa: 0.2280, the gradient boosting machine model is the best performance model.

6. The method for predicting brain metastasis in patients with non-small cell lung cancer according to claim 5, characterized in that: A visualization tool was established based on the existing model, and the patient's clinical information was input to predict the patient's risk of brain metastasis.

7. A brain metastasis prediction system for patients with non-small cell lung cancer, characterized in that: include: A data acquisition module, which acquires patient data based on the SEER database, wherein the patient data includes a set of variables of patient baseline information; Data preprocessing module, excluding variables with missing values ​​and using LASSO algorithm to remove patient data with insignificant variables; The model building module randomly divides the screened patient data into training sets, validation sets, and test sets, and builds multiple machine learning models to achieve the best prediction performance; The statistical analysis module trains the machine learning model based on the training set and the validation set, uses the test set and evaluates the above training results based on the ROC curve area, calibration curve, and model performance parameter evaluation model to select the model with the best performance.

8. A storage device, characterized in that: The storage medium stores a plurality of instructions, which are suitable for being loaded by a processor to execute the steps in the method for evaluating the cardiotoxicity of immune checkpoint inhibitors according to any one of claims 1 to 6.

Images