Risk prediction system for esophageal squamous cell carcinoma based on clinical phenotype and logistic regression analysis

Abstract

The invention provides a risk prediction method for esophageal squamous cell carcinoma based on clinical phenotype and logistic regression analysis, which is used to realize the prognosis and survival risk assessment of patients with esophageal squamous cell carcinoma. The steps are as follows: first, according to the clinical detection data of esophageal squamous cell carcinoma patients, the characteristic indicators are screened, and a decision tree classifier is constructed according to the characteristic indicators; secondly, the esophageal squamous cell carcinoma patients are divided into early stage and middle and late stage esophageal squamous cell carcinoma by using the decision tree classifier. Then, the blood index information of patients with esophageal squamous cell carcinoma one week before surgery was obtained, and the blood indexes with high correlation with the survival risk of patients with esophageal squamous cell carcinoma were screened out and a logistic regression model was constructed; The indicators were input into the logistic regression model to obtain the probability value of the prognostic survival risk of patients with esophageal squamous cell carcinoma, and then to determine the level of the prognostic survival risk. The invention more accurately judges the survival state of patients with esophageal squamous cell carcinoma after operation, improves the performance of risk prediction, and reduces the cost of risk prediction.

Description

technical field [0001] The invention relates to the technical field of machine learning, in particular to a risk prediction method for esophageal squamous cell carcinoma based on clinical phenotype and logistic regression analysis. Background technique [0002] With the increasing incidence of cancer, model-based prediction of cancer prognosis has been widely used in different diseases, and accurate prognosis of cancer patients is still the primary problem currently faced. The clinically detected data is typically characterized by multicollinearity, high dimensionality, and high noise, which makes the data itself have problems such as information redundancy and nonlinearity. The problem is that on the one hand, the "high dimension" makes the processing of data require high computational costs, and on the other hand, the data itself cannot directly reflect the essential properties. In recent years, scholars at home and abroad have considered and discussed the issue of dimens...

AI Technical Summary

Problems solved by technology

[0004] In view of the deficiencies in the above-mentioned background technology, the present invention proposes a risk prediction method for esophageal squamous cell carcinoma based on clinical phenotype and logistic regression analysis, which solves the technical problem of low recognition rate caused by incomplete feature screening of existing prediction models

Images