Medical Data Mining And Appling Research Based On The Deep Learning

With the development of the medical equipment and technology,we are brought into the age of medical big data.Information hidden in the medical data is helpful for disease prevention,diagnosis and treatment,It's difficult to use the traditional medical technology to find value and accurate information from various and complex data,so we need to develop existing technology to solve disease problems from various data for people.Data mining technology provides a new way to achieve the effective use of medical information resources.Cancer treatment is a big problem,screening and diagnosis are the most effective ways to prevent cancer,so this thesis designs medical data mining schemes of cancer screening and diagnosis based on the deep learning technology,which provide a scientific basis for cancer screening and diagnosis.First,this thesis uses the artificial neural network learning from the physical examination data for cancer screening.Because of complicated steps and high cost of cancer detection,this thesis with the goal to simplify the detection achieves cancer screening from physical examination data.The model in this thesis improves the positive rate through the ROC curve and proved that the feasibility of cancer screening based on physical examination data.Second,according to the Wisconsin Breast Cancer Dataset(breast cancer cell tissue section),this thesis changes the typical two-dimensional convolution neural network structure into a one dimensional structure to adapt to the input data and achieved breast cancer diagnosis.Compared with other algorithms,the results show that the convolutional neural network in this thesis has better sensitivity,specificity,accuracy and the ROC curve.It also verifies the convolutional neural network can not only be applied to handle unstructured data in image recognition field,but also handle structured medical data.This thesis studies cancer screening and diagnosis model based on deep learning technology,which provides a reference for the early prevention and early diagnosis of cancer.