The Design And Implementation Of A Machine Vision-based Pulmonary Symptom Assisted Diagnosis System

In the cancer statistical report released in 2018,lung cancer accounts for 11.6% of the total number of cancers,ranking first in the world,especially in China.And in cancer,there is essential difference between benign cancer and malignant cancer.If lung cancer can be found and diagnosed in the early examination,it is very beneficial for the follow-up diagnosis and treatment,but the accuracy of diagnosis depends on the experience of doctors.Moreover,there are a large number of clinical trial data in the medical system,doctors can use these data as a reference to give the most appropriate At this time,it is necessary to quickly obtain the most similar clinical trial information.In clinical diagnosis,the use of medical image diagnosis is an essential step in the diagnosis of lung cancer.The common methods include computed tomography(CT)scanning.Through CT scan,doctors can judge whether a patient may have lung cancer by whether there is a nodule in the lung.If a nodule is found,the final diagnosis of lung cancer needs to be made by means of pathological examination.In the detection of pulmonary nodules,the convolution neural network technology is used to classify the lung images input by doctors to detect whether there is pulmonary nodules.After detecting the pulmonary nodules,case examination is needed.At this time,there are many means for pathological examination.Image matching method is used to obtain similar pictures and corresponding pathological examination steps of patients,which helps doctors to give the examination method Methods: after pathological examination,if lung cancer is confirmed,clinical diagnosis and treatment should be carried out according to the results of pathological examination.At this time,pathological information matching can be carried out to obtain the diagnosis and treatment process and clinical trial information of patients with the most similar pathological information,which can be used as a reference for subsequent diagnosis and treatment of patients.Through the combination of image classification,image matching and information matching,it can help doctors in the detection of pulmonary nodules,pathological examination of lung cancer and diagnosis and treatment means after diagnosis,and the accuracy rate in the process of image classification has reached 90%.Image matching and case information matching can get the best results and display accurately.