Construction method of imaging omics model for pancreatic cancer prediction

Abstract

The invention discloses a construction method of an imaging omics model for pancreatic cancer prediction. The construction method provided by the invention comprises the following steps: acquiring data; feature extraction; constructing a feature dimension reduction and image omics model; constructing and verifying an image omics model; survival analysis. The method has the following beneficial effects that the method has good practicability; the imaging omics model for pancreatic cancer early stage and progression stage classification and prognosis prediction constructed by the method can accurately distinguish pancreatic cancer patients in the early stage and progression stage, predict prognosis of the patients, analyze all focus tissues instead of being limited to local parts (such as puncture and the like), and provide more comprehensive information of focuses. The method is non-invasive, non-invasive, good in repeatability and comprehensive in detection, does not increase extra medical expenses of patients, and is easy to operate and convenient for clinical popularization and application.

Description

technical field [0001] The invention relates to the technical field of pancreatic cancer detection, in particular to a method for constructing a radiomics model for clinical staging and prognosis prediction of pancreatic cancer. Background technique [0002] Pancreatic cancer is one of the most malignant tumors with a very poor prognosis, with a 5-year survival rate of only 9.3%. At present, radical surgical resection is the most effective method for the treatment of pancreatic cancer. However, due to the lack of typical symptoms and signs in the early stage of pancreatic cancer, more than 80% of pancreatic cancer patients lose the opportunity of surgical resection due to advanced lesions when they see a doctor. Therefore, the early detection, accurate staging and surgical resectability assessment of pancreatic cancer, so as to adopt a reasonable and scientific comprehensive treatment is the key to improve the survival rate of patients. At present, TNM staging of pancreatic...

AI Technical Summary

Problems solved by technology

[0007] The purpose of the present invention is to provide a method for constructing a radiomics model for pancreatic cancer prediction, to solve the problem that the existing detection method of endoscopic early pancreatic cancer imaging uses invasive detection to cause discomfort to patients, and to prepare targeted nano The particle operation steps are relatively complicated, and the cost of manpower and material resources is high

[0009] In order to realize the construction method of the above-mentioned radiomics model for pancreatic cancer prediction, the existing detection method of endoscopic early pancreatic cancer imaging uses invasive detection to cause discomfort to patients, and the operation steps of preparing targeted nanoparticles are relatively complicated and labor-intensive. The problem of higher material cost, the present invention provides the following technical solutions:

Images