Rough Set Theory Application Of Computer Aided Diagnosis In Lung Cancer

Background Lung cancer is the cancer with the highest morbidity and mortality in the world,which seriously threatens human life and health.Early diagnosis and treatment of lung cancer can fundamentally improve the survival rate of lung cancer patients,and even cure lung cancer.Imaging images are an important reference basis for staging diagnosis of lung cancer.With the explosive growth of medical imagery,doctors have a high rate of missed diagnosis and low accuracy.Computer-aided diagnosis based on medical imaging images can help improve the sensitivity and specificity of doctors' diagnosis.Attribute reduction is a vital method to obtain valuable information from data.It can reduce the dimension of characteristic attributes and simplify the processing of knowledge.Rough set theory is a mathematical tool that deals with inaccurate,inconsistent,and incomplete information.Only using the information provided by the data itself,rough set-based attribute reduction can find the rules of the problem and carry out attribute reduction without prior knowledge in data mining.Therefore,the rough set-based attribute reduction algorithm has been used as an important tool for building computer-aided diagnostic models because of its precise analytical processing power.Objectives Using medical imaging images of lung tumors with doctor's orders(CT,PET,PET \ CT,3000 cases each)as research data,a computer-aided diagnosis model of lung tumors based on rough set theory is built to improve the diagnosis accuracy of doctors and reduce the misdiagnosis rate of doctors' diagnosis.Methods First,establishing a computer-aided diagnosis model of lung tumor based on rough set theory to solve the problem of high rate of false positives in the computer-aided diagnosis system of lung tumor.Second,building computer-aided diagnosis model of lung tumor imaging based on integrated DE-NRS to solve the problem of poor fault tolerance of classical rough set theory and its inability to handle continuous data.Finally,aiming at the construction problem of the performance of differential evolution algorithm depending on the of control parameters(variation coefficient(F),crossover coefficient(CR))and fitness function,a high-dimensional attribute reduction model of lung tumor based on JADE-NRS is built.Results The computer-aided diagnosis model of lung tumor imaging omics based on integrated DE-NRS has good overall performance in the identification of benign and malignant lung tumors,and the recognition accuracy reaches 99.72%.It has good robustness and scalability.Compared with the DE-NRS model,the JADE-NRS model performs attribute reduction on CT medical image images of high-dimensional lung tumors.The obtained attribute subsets are identified by the SVM classifier,and the recognition accuracy ACC is increased by 1.69%,and the time is shortened by 57.7658 s.