161 results about "Logistische regression" patented technology

The invention discloses a cardiovascular disease risk prediction method and system based on dietary multi-modal data and ensemble learning, and the method comprises the steps: constructing a multi-modal set through integrating multi-source heterogeneous data such as demographic statistics, dietary nutrition, clinical physiological and biochemical indexes and lifestyles; data cleaning is completed based on a box plot method and missing value processing, and key features are screened through Pearson's correlation coefficients, variance expansion factors and feature importance evaluation; a plurality of heterogeneous base models are fused by adopting a Stacking integration framework, a meta-feature matrix is generated through five-fold layered cross validation, and multi-level decision fusion is realized through a logistic regression meta-model; and quantifying the contribution weight of the dietary characteristics to the risk in combination with an SHAP method, and generating a visual interpretation chart and personalized intervention suggestions. According to the method, the accuracy and the stability of a prediction result are remarkably improved, the contribution degree and the action mechanism of dietary factors and other characteristics to the prediction result can be deeply analyzed, powerful support is provided for accurate prevention and personalized treatment of cardiovascular diseases, and the method has good practical value.

The invention relates to the field of noninvasive prenatal diagnosis, and particularly discloses a preeclampsia noninvasive screening method based on deep sequencing 8bp oligonucleotide double-fragment characteristics, which comprises the following steps: collecting preeclampsia and healthy pregnant woman peripheral blood samples, and extracting free DNA for high-throughput sequencing; the method comprises the following steps: extracting core 8-mer sequences' GTGCGCCC 'and' GATGGGGT 'in a long fragment of 150-200bp through bioinformatics analysis; an integrated support vector machine, K-nearest neighbor, extreme gradient lifting, a random forest and a multi-layer perceptron are combined with a logistic regression element classifier to construct a stacking model, the frequency of a core sequence is normalized, machine learning analysis is carried out, and the preeclampsia risk is predicted. According to the invention, two 8bp oligonucleotide characteristic fragments are specifically screened, and a deep learning architecture of multi-model fusion is combined, so that the limitations of low specificity and invasive detection of a traditional screening method are effectively broken through.

The invention discloses a pelvic floor assessment method and system combining pelvic floor myoelectricity and muscle force signals, and relates to the technical field of pelvic floor muscle assessment, and the method comprises the steps: synchronously collecting pelvic floor myoelectricity signals and pressure signals, and recording data of a pre-resting stage, a rapid contraction stage, a tension contraction stage, an endurance contraction stage and a post-resting stage; the pelvic floor electromyographic signals and the pressure signals are preprocessed, and statistical characteristics are extracted based on the preprocessed signals; constructing a pelvic floor evaluation model, wherein the pelvic floor evaluation model comprises a logistic regression model, a long-short term memory network based on a cross attention mechanism and a multi-layer perceptron classifier; and training the pelvic floor evaluation model by using an Adam optimizer and a cross entropy loss function, and optimizing parameters of the pelvic floor evaluation model through evaluation indexes. According to the pelvic floor evaluation model, the timing sequence features and the statistical features are combined, more comprehensive signal interpretation is achieved, the generalization ability of the pelvic floor evaluation model is improved, and the pelvic floor health condition can be recognized more accurately.

The invention provides a diabetes chronic complication early warning method and device, a medium and computer equipment. The early warning method comprises the following steps: acquiring body data of a diabetic patient; screening the body data by adopting an LASSO logistic regression model to generate a data training set; training the multi-layer perception model by using the data training set to obtain a diabetic complication early warning model; obtaining to-be-detected body data, inputting the body data into the diabetic complication early warning model for prediction after screening, and generating an early warning result; by the adoption of the method, dimension reduction is conducted on the collected body data through the LASSO logistic regression model, the training data of the model can be effectively reduced, the computing power requirement for the training and reasoning process is lowered, the model can be conveniently deployed to a local computing end, and therefore the diabetic patient can conveniently achieve home detection. Compared with conventional blood detection, the non-invasive blood detection device has the advantages of being non-invasive and small in operation difficulty.

The invention relates to a training method of a respiratory tract infection disease progress and prognosis prediction model. The training method comprises the following steps: extracting mRNA from peripheral blood of a target patient, and carrying out transcriptome sequencing to obtain a sequencing result; based on the ferroptosis related gene set, comparing ferroptosis score differences of two groups of patients with community-acquired pneumonia and sepsis, and screening corresponding ferroptosis related genes with statistical significance from a sequencing result; screening out genes meeting preset conditions from the ferroptosis related genes based on LASSO regression; and establishing an RTI clinical outcome prediction model through logistic regression by taking whether the patient is sepsis or not as an outcome dichotomy variable and taking the screened gene expression quantity as a prediction variable. According to the invention, after the prediction model is subjected to machine learning screening such as LASSO and the like, the core feature with the highest prediction value is reserved, so that the risk of over-fitting of the model on training data is reduced.

The invention relates to the technical field of biological medicine, in particular to construction and application of a candida patient death risk prediction model. The model is based on traditional SOFA scoring, core variables are screened through single-factor and multi-factor logistic analysis and LASSO regression analysis, respiratory system, blood coagulation function and circulatory system SOFA scoring and key indexes such as the maximum lactic acid value, the minimum albumin value and the maximum blood urea nitrogen (BUN) are integrated, and a prediction model is constructed through logistic regression. Through multi-center data verification, the AUC of an internal verification queue reaches 0.826, the AUC of an external verification queue is 0.813, and a calibration curve shows that the predicted death rate is highly consistent with the actual death rate. Patients can be divided into a high risk group and a low risk group through a 13-score threshold value, and the 28-day survival rate difference is significant (plt; 0.05) of the method. An efficient and convenient early risk assessment tool is provided for critical candida patients, and clinical precise intervention and prognosis improvement are assisted.

The method for monitoring parameters of a patient during surgery with extracorporeal circulation serves to estimate the presence of AKI risk and comprises calculating a dynamic global index of AKI risk as a logistic regression deriving from a static AKI risk index and a dynamic index of AKI risk. The value and / or the time trend of such global index of AKI risk can be advantageously displayed on a monitor during the surgery. The dynamic AKI risk index is calculated at least on the basis of the extracorporeal circulation time, the minimum level of oxygen supply and the exposure time to oxygen supply below a critical threshold which are measured or determined repeatedly during surgery. Preferably, the dynamic AKI risk index is also calculated on the basis of the minimum mean arterial pressure during surgery and / or the maximum concentration of lactate in the blood during surgery and / or the minimum hematocrit during surgery and / or the fact that the patient has been subjected to transfusion(s) during the extracorporeal circulation.

The invention relates to a disease information management method and system based on multi-source data, and relates to the technical field of medical artificial intelligence, and the method comprises the steps: forming multi-source heterogeneous data through employing the obtained infectious disease data and chronic disease data, and carrying out the preprocessing with the corresponding data features as a reference, on one hand, risk prediction and clustering analysis are performed on infectious disease related data through a logistic regression model to obtain risk assessment information, and a propagation trend is predicted through the model; on the other hand, chronic disease data are predicted and evaluated through the model, a medical advice result is obtained, and finally disease control information management is conducted in combination with all the data. Therefore, a set of perfect disease prevention and control information management is realized, on one hand, more timely and accurate infectious disease detection and early warning are realized, and the efficiency and effect of major infectious disease prevention and control work are improved, and on the other hand, chronic disease management is perfected, and the problems of diagnosis and treatment errors and lagging of patient treatment consciousness caused by incomplete data are avoided.

The invention discloses an early-stage lung cancer prediction method based on a multi-modal eccDNA marker, and relates to the technical field of liquid biopsy, and the method comprises the following steps: collecting a peripheral blood sample, carrying out differential centrifugal separation on plasma, constructing a cfDNA library, and carrying out double-end sequencing to obtain original sequencing data; extracting structural features through an eccDNA model, and inputting the structural features into a first machine learning model to generate a circular DNA score; extracting variation features through an SNV model, and inputting the variation features into a second machine learning model to generate a spectrum feature score; extracting copy number features through a CNV model, and inputting the copy number features into a third machine learning model to generate copy number scores; carrying out probability distribution calibration on the circular DNA score, the spectrum feature score and the copy number score; inputting the three types of molecular features into a deep neural network to generate a first fusion score; inputting the calibrated score into a logistic regression model to generate a second fusion score; and generating a final lung cancer risk probability according to the first fusion score and the second fusion score.

The application discloses an evaluation method for predicting the safety of EGFR TKIs monotherapy for NSCLC. The method comprises the following steps: obtaining the data of a plurality of patients receiving EGFR TKIs treatment; taking MDRAE (maximum grade of drug-related adverse event) as a safety index, constructing a safety evaluation final model, and the final model is an ordered logistic regression model, the covariates of the final model include normalized steady-state trough concentration, EGFR TKIs treatment history, gender, baseline glutamyltransferase, baseline uric acid, baseline creatine kinase, baseline platelets and baseline lymphocytes; obtaining the data of a target patient, inputting the numerical values of each covariate in the data of the target patient into the final model; determining the first probability of MDRAE of the target patient being grade 3 and above, the second probability of MDRAE being grade 2 and the third probability of MDRAE being grade 1 and below according to the final model, and taking the grade corresponding to the maximum probability in the three probabilities as the predicted grade of MDRAE of the target patient. The application improves the accuracy of predicting the safety of EGFR TKIs monotherapy for NSCLC.

The invention relates to a novel marker for a glioblastoma tumor transition region and application of the novel marker, and belongs to the technical field of biological detection. In order to solve the technical problems that in the prior art, tumor invasion leading edge markers cannot be effectively recognized, and postoperative residues and prognosis are difficult to accurately evaluate, 19 gene markers with high specificity expression in a tumor transition region, such as SELENOP, SHTN1 and CTNNA3, are screened out by combining space transcriptome analysis with a machine learning algorithm. The expression level of the marker is detected through digital imaging, nucleic acid sequencing and other technologies, a logistic regression prediction model with the AUC value larger than 0.95 is constructed, and the logistic regression prediction model is applied to diagnostic kit preparation, tumor layering screening and prognosis evaluation. According to the invention, precise recognition of the tumor transition region is realized, and a molecular basis is provided for postoperative recurrence monitoring, individualized treatment scheme making and novel treatment target screening.

The invention relates to the technical field of medical imaging, in particular to a breast cancer tissue HER2 automatic scoring method and device based on a characteristic curve, and the method comprises the steps: carrying out the reading operation of a full-slice image of immunohistochemical staining; extracting a target region of interest of the full slice image; slicing and screening the target region of interest to obtain a plurality of image slices; calculating a percentage-saturation characteristic curve of each image slice, and extracting characteristic information of the percentage-saturation characteristic curve; based on the feature information of the percentage-saturation feature curve, performing preliminary HER2 scoring on each image slice in combination with a rule base and a logistic regression classification algorithm; and integrating the scoring results of the image slices, and outputting a final HER2 score and a characteristic curve visualization graph. According to the method, the feature curve is combined with the rule base and the logistic regression classification algorithm, so that the objectivity of HER2 scoring is improved, and the problems of subjectivity and low efficiency of manual evaluation are reduced.

The invention discloses a neonatal congenital heart disease screening method based on machine learning, and the method comprises the following steps: 1, designing a questionnaire according to literatures related to adverse birth result risk factors, collecting infant cardiac ultrasound examination data, and then preprocessing the data; step 2, randomly dividing the data obtained in the step 1 into a training set and a test set, and performing dimension reduction processing on features by using lASSO regression; and step 3, based on the features obtained in the step 2, training a prediction model by using machine learning algorithms such as a random forest, a support vector machine, a lightweight gradient elevator, logistic regression and extreme gradient lifting. According to the method, the model is optimized through combination of multi-step data processing and various machine learning algorithms, related features can be comprehensively and scientifically screened, the accuracy of the prediction model is improved, a more reliable and effective method is provided for screening the neonatal congenital heart disease, early-stage accurate screening is facilitated, and the screening efficiency and quality are improved.

The invention discloses an ADHD auxiliary diagnosis model construction method, control equipment and a program product, relates to the field of medical artificial intelligence, and obtains a high-performance ADHD auxiliary diagnosis model through a small sample. The method comprises the following steps: preprocessing acquired sample data, screening out a predetermined number of indexes with the importance higher than that of a diagnosis result in the sample data to obtain a sample set, and dividing the sample set to obtain a training set and a test set; and constructing a logistic regression classifier model, training the logistic regression classifier model by using the training set, and testing the trained logistic regression classifier model by using the test set to obtain the ADHD auxiliary diagnosis model. According to the method, the problem that an ADHD prediction model depends on the sample size is solved, and the feasibility of clinical application is improved.

The invention discloses a system and a method for analyzing and predicting malnutrition of patients with elderly mild cognitive impairment, belongs to the technical field of malnutrition analysis and prediction, incorporates a more comprehensive and systematic index system, and covers potential risk factors, traditional risk factors and laboratory data of the patients with elderly mild cognitive impairment. The comprehensive breakthrough of the research perspective is realized, and the limitation defect of the prior art is broken through. The decision tree model and the logistic regression model are comprehensively compared, a precise malnutrition risk prediction model for the elderly MCI patient is constructed, a reference with practical value is provided for clinical prevention and treatment work, and optimization and perfection of a malnutrition prevention and treatment strategy of the elderly mild cognitive impairment patient are expected to be promoted. According to the method, the defects that in the prior art, analysis of the malnutrition of the elderly mild cognitive impairment patient has limitation, and a traditional prediction model is insufficient in decision making and pertinence aspects of the malnutrition of the elderly mild cognitive impairment patient are effectively overcome.

The invention provides a privacy protection calculation method based on group learning and cardiovascular disease prediction, and the method comprises the steps: obtaining local medical data, carrying out the preprocessing, and dividing the data into a training set and a test set; training at least one base learner according to the local data characteristics and outputting a prediction probability; a Gaussian difference privacy mechanism is adopted to disturb prediction output; dynamically adjusting the privacy budget epsilon in an exponential decay mode according to the number of rounds in the training process; receiving prediction results after disturbance of all the clients and aggregating the prediction results by using a logistic regression model as a meta-learner; training a lightweight student model through knowledge distillation; the prediction capability is evaluated according to multiple performance indexes, and the influence degree of the differential privacy mechanism on the model performance is verified through a model test result under multiple epsilon values; and constructing an attack model to evaluate the protection effect of the privacy mechanism. On the premise of ensuring data privacy security, the cardiovascular disease prediction capability is improved, and the optimal balance between privacy protection and model performance is realized.

The invention relates to the technical field of bioinformatics and medical data processing, and discloses a polygene molecular diagnosis model and a construction method and application thereof.The construction method comprises the steps that a data acquisition module constructs a genome, clinical phenotype and environmental exposure data matrix, a preprocessing module executes regression filling and standardizes continuous variables, and a data processing module performs data processing; the feature screening module executes double-layer screening by using LASSO and a random forest model to output a core feature subset, the model building module builds a logic regression architecture to calculate a baseline logarithm probability, and the dynamic updating module outputs a real-time risk probability in combination with follow-up visit environment data, a time adjustment coefficient and the baseline logarithm probability. And the interactive output module outputs a risk layering label and a feature contribution degree. According to the method, redundancy is eliminated through double-layer screening, the time dimension is introduced to adjust the real-time correction probability, visual attribution is realized in combination with the SHAP algorithm, and the dynamic monitoring capability and interpretability are improved.

This paper provides a method and system for constructing a two-stage cancer survival prediction model. In the first stage, a logistic regression classifier based on SVM (SMOTE) is used to predict whether patients will survive five years in the first stage. In the second stage, a random forest regression model is used to predict the specific survival time (in months) for patients who are not expected to survive five years. This research enables effective and targeted predictions based on unique patient data, helping to improve treatment outcomes and patient quality of life.

The application discloses a classification method and device for pancreatic neuroendocrine tumors. The classification method comprises the following steps: segmenting tumor masks from enhanced CT images; for each tumor mask, generating a first peritumoral region and a second peritumoral region; dividing the tumor into multiple habitats for each tumor mask; extracting 3D radiomics features, multiple peritumoral microenvironment features, multiple tumor habitat features, multiple local pathological features and multiple global context features; selecting multiple target features for each enhanced CT image and calculating a radiomics score; inputting the radiomics score and clinical variables into a logistic regression model for training to obtain a trained logistic regression model; and using the trained logistic regression model to classify the enhanced CT images to be classified. The application can quickly and accurately realize G-grade classification of pancreatic neuroendocrine tumors, and provides a reliable non-invasive evaluation tool for clinical decision-making.

The invention discloses a marker combination, a kit thereof and application of the marker combination to prediction of diabetic peripheral neuropathy, the marker combination comprises human angiotensin (ANG), human vascular endothelial cell adhesion molecule 1 (VCAM1) and human mannan binding lectin serine peptide 1 (MASP1), TMT quantitative proteomics and PRM targeted proteomics are adopted, and the marker combination is used for detecting diabetic peripheral neuropathy. A technology combination strategy of an ELISA experiment is adopted, and key differential expression proteins such as MASP1, VCAM1 and ANG in serum exosomes of patients with diabetic peripheral neuropathy are systematically screened and verified. By establishing a multi-model machine learning diagnosis system based on logistic regression, a support vector machine and a naive Bayes algorithm, simple and rapid identification of diabetic peripheral neuropathy patients is realized, and a new experimental basis is provided for clarification of a molecular regulation network of diabetic peripheral neuropathy; more importantly, a rapid screening tool with clinical application potential is developed.

This invention relates to an MRI-based method for classifying benign and malignant intraductal lesions (IDLs) of the breast, comprising the following steps: Step 1) Extracting a set of clinical features from MRI images: age, ADC value, BI-RADS category, lesion shape, margin clarity, and TIC curve type; Step 2) Extracting radiomics features from the 3D lesion segments of the T1 DCE-MRI sequence; Step 3) Reducing the dimensionality of the radiomics features using the Lasso algorithm; Step 4) Fusing the dimensionality-reduced radiomics features with the clinical features into an input vector; Step 5) Outputting benign / malignant probability values ​​through a logistic regression model; Step 6) Applying stratification rules to BI-RADS 4A lesions: older patients or those with low ADC values ​​are marked as high-risk and biopsy is recommended; younger patients with high ADC values ​​are marked as low-risk and follow-up is recommended. This method solves the problem of current dynamic contrast-enhanced MRI's difficulty in accurately identifying BI-RADS categories, especially the BI-RADS 4A high-risk subgroup, thereby reducing missed diagnoses and over-biopsy. It is of urgent significance for overcoming the bottleneck of preoperative risk stratification in IPLs and guiding individualized clinical decision-making.

This invention discloses a method for predicting sudden cardiac death based on forensic autopsy data using machine learning and its forensic applications, belonging to the fields of machine learning, statistics, and forensic identification. This invention provides a method for screening independent predictors of forensic diagnosis of sudden cardiac death using LASSO regression and logistic regression. Using 10-fold cross-validation, the method selects the minimum lambda(λ) to identify 14 risk factors. Logistic regression ultimately identifies 9 independent predictors, including age, heart weight, left ventricular wall thickness, right ventricular wall thickness, interventricular septum thickness, aortic valve circumference, mitral valve circumference, liver weight, and left kidney weight. A nomogram and a web-based calculator are constructed for predicting sudden cardiac death in forensic practice, thereby determining sudden cardiac death using objective indicators. This method has not been reported worldwide.

The invention specifically discloses a medication safety assessment method and system for hypertensive patients, and the method comprises the following steps: extracting one-complaint five-history, medication records, inspection indexes and outcome data of the hypertensive patients from an electronic health record system (EHR), carrying out the preprocessing, extracting key entities, and carrying out the verification of the key entities; constructing a comprehensive data set of'co-disease-medication-examination-symptom 'of the patient, identifying a co-disease combination strongest associated with hypertension, and mining specific adverse reaction signals under the co-disease combination; key factors causing adverse reactions or unreasonable medication are attributed through a multi-factor conditional logic regression or decision tree model; and based on common disease combination and multi-factor logistic regression analysis of one complaint and five histories, performing pairing t inspection on laboratory indexes, and generating a medication safety report of the hypertension common disease patient. By adopting the technical scheme, through multi-dimensional data analysis, specific influence factors causing adverse drug reaction or unreasonable drug use are attributed, and accurate drug use safety guidance is provided for clinicians.

This invention provides a method for analyzing the radiomics characteristics of medullary thyroid carcinoma based on ultrasound, relating to the field of image or video recognition or understanding technology. The method includes: obtaining data from thyroid surgeries at various target hospitals to form a backup dataset; filtering the backup dataset to obtain a target dataset; labeling the data in the target dataset with Regions of Interest (ROIs); performing feature extraction and feature filtering on the data in the target dataset to obtain training and validation sets; constructing a target radiomics feature model; calculating radiomics feature scores based on the target radiomics feature model; obtaining a clinical feature model, an ultrasound feature model, a comprehensive model, and a scoring model based on logistic regression analysis; and performing radiomics feature analysis based on the clinical feature model, ultrasound feature model, comprehensive model, and scoring model. This invention solves the problems of low diagnostic accuracy and high requirements for experience of image interpreters in existing technologies for thyroid nodules.

The application provides a breast cancer patient lymph node metastasis prediction model without clinical pathological characteristics, comprising the following steps: after three-dimensional reconstruction of two-dimensional lung enhanced CT films, an axillary lymph node atlas is established, all axillary lymph nodes in the atlas are selected as ROI regions, and more than 5 combined image features of each axillary lymph node are selected to distinguish whether breast cancer has axillary lymph node metastasis; and a logistic regression machine learning prediction model is used to construct the breast cancer patient axillary lymph node metastasis prediction model. The model established by the application can non-invasively predict whether breast cancer has axillary lymph node metastasis, the clinical pathological characteristics of the patient are not included in the model, and the image cutting in the model is not based on breast tumors, but based on axillary lymph nodes; the model is used to determine a suitable axillary treatment scheme, thereby avoiding unnecessary axillary surgery and complications, and helping to carry out more accurate surgery and adjuvant therapy mode of breast cancer.

This invention discloses a method for identifying intrahepatic cholangiocarcinoma (ICC) based on serum peptide characteristics and its application. The invention collects serum samples from three groups: individuals with ICC, those with benign liver disease, and healthy controls. Peptide characteristic peaks associated with ICC are screened, and a model is constructed using a logistic regression algorithm. This model demonstrates excellent discriminative ability on both the training and independent test sets, with AUCs reaching 0.986 and 0.963, respectively, significantly outperforming traditional tumor markers CA19-9 and CEA. The peptide characteristic peak combination and the discriminative model constructed based on it can be used for early detection of ICC, differentiation between benign and malignant cases, and screening of high-risk populations. It offers advantages such as rapid detection, standardized procedures, strong early ICC identification capability, and low false positive rate, providing a highly sensitive, specific, and widely applicable non-invasive screening tool for clinical use.

The invention relates to application of a reagent for evaluating the recurrence risk of epithelial ovarian cancer, and belongs to the technical field of molecular biology. The application specifically refers to application of a reagent for specifically detecting the expression level of SNORD18C and / or SNORD84 in an epithelial ovarian cancer operation tissue sample in preparation of a kit for evaluating the recurrence risk of an epithelial ovarian cancer patient subjected to resection. The change of the expression quantity of SNORD18C and SNORD84 in an epithelial ovarian cancer operation tissue sample is found, and the change of the expression quantity is closely related to the recurrence risk of an epithelial ovarian cancer patient after the epithelial ovarian cancer patient is subjected to resection. Therefore, by detecting the expression of SNORD18C and / or SNORD84 in an epithelial ovarian cancer tissue sample and combining the binomial logistic regression model, the recurrence risk of an epithelial ovarian cancer patient after the resection is assessed, and high sensitivity and specificity are achieved.

We describe a computer-implemented method for determining a patient's brain age and optionally stratifying patients into dementia risk groups based on the determined brain age. The methods comprise extracting at least one volumetric feature from an image of a brain by: obtaining at least one volume value for at least part of the patient's brain, and normalising the at least one obtained volume value to obtain the at least one volumetric feature. Brain age is predicted by inputting the at least one extracted volumetric feature into a pre-trained brain age model, wherein the brain age model is a linear regression model. Bias of the linear regression model may also be corrected. A classification model, such as a logistic regression binary classifier, may be used to stratify patients into dementia risk groups.

The invention discloses a method for constructing a prediction model for subcutaneous fat hyperplasia caused by insulin injection. The method comprises the following steps: S1, selecting a plurality of participants for questionnaire survey according to inclusion standards; s2, performing ultrasonic inspection on all participants; s3, performing feature screening on the predicted variables by adopting LASSO regression; s4, developing a fat hyperplasia prediction model by using three algorithms of random forest, extreme gradient lifting and logistic regression, and selecting the fat hyperplasia prediction model with the optimal performance for research; and S5, explaining the influence of the predictive variables on the fat hyperplasia prediction model through an SHAP method. According to the method, a machine learning-based insulin injection-induced fat hyperplasia prediction model is created, the constructed extreme gradient lifting machine learning model shows relatively high efficiency in the aspect of predicting the occurrence of fat hyperplasia of the diabetic patient, and the diabetic patient with high-risk fat hyperplasia can be accurately identified.

The invention discloses an evaluation method for predicting the safety of EGFR TKIs (epidermal growth factor receptor TKIs) single-drug treatment on NSCLC (non-small cell carcinoma). The method comprises the following steps: acquiring data of a plurality of patients receiving EGFR TKIs treatment; the method comprises the following steps of: constructing a safety assessment final model by taking MDRAE (Maximum Grade Of Related Adverse Event) as a safety index, the final model is an ordered logic regression model, and the covariables of the final model comprise normalized steady-state valley concentration, EGFR TKIs treatment history, gender, baseline glutamyltransferase, baseline uric acid, baseline creatine kinase, baseline platelet and baseline lymphocyte; obtaining target patient data, and inputting each covariable value in the target patient data into the final model; and according to the final model, determining a first probability that the MDRAE of the target patient is at a level 3 or above, a second probability that the MDRAE is at a level 2, and a third probability that the MDRAE is at a level 1 or below, and taking a level corresponding to a maximum probability value in the three probabilities as an MDRAE prediction level of the target patient. According to the invention, the accuracy of predicting the safety of EGFR TKIs single-drug treatment on NSCLC is improved.