Logistic Regression And Neural Network Method In The Application Of The Medical Diagnosis

As one of the most common malignant tumor,colorectal cancer makes a serious threat to people's lives.It is now mainly treated by operation and lymph node metastasis is one of the main influencing factors of prognosis.In this paper,statistical methods were used to explore the metastasis rules and influencing factors of colorectal cancer,which would be guidance for operation treatment and reduce the incidence of postoperative recurrence.By analyzing the conventional indicators of 252 patients with colorectal cancer from the people's Hospital of Liaoning Province,this paper applied three methods named logistic regression analysis,neural network and SCAD method to select the main influencing factors of lymph node metastasis,followed by comparisons of these methods.The first part of this paper described the current status on the research of colorectal cancer,along with the background and significance.The second part introduced the relevant theoretical knowledge.The third part was a specific analysis based on the data of patients with colorectal cancer from the People's Hospital of Liaoning Province.The fourth part was the conclusions of this paper and its weaknesses.