| In the early 2020 s,the 2019-ncov,which shocked the world,caused serious damage to the world’s human health and caused huge losses to the world economy.According to statistics,Lung Disease is one of the diseases with the highest morbidity and mortality in the world,and lung cancer has become the first malignant tumor among all cancers because of its most difficult to detect,high recurrence and difficult to cure.Therefore,early detection,early screening,early treatment in the prevention of Epidemic Diseases,cancer play an important role in the spread of patients can improve the survival rate and cure rate,to minimize the loss.With the development of artificial intelligence and computer-aided diagnosis system,the traditional methods of lung cancer detection in CT images not only rely on the experience of high-level medical staff,but also have complicated steps and low diagnostic accuracy.So this paper presents a lung nodule detection system of CT image based on Yolo V3 Algorithm.In this paper,an automatic detection method of lung nodules based on the object detection algorithm of regression is proposed,and a set of lung CT image processing flow which combines the threshold segmentation algorithm with the digital morphological processing is presented,we also trained and studied 1,186 lung nodules from LUNA16 data set,which is the authoritative data set for Lung Disease Research,and observed Yolo V3 evaluation results in the data set to validate the model,at the same time,a computer aided diagnosis system of CT pulmonary nodules is designed based on CAD technology.The experimental results show that Yolo V3 model can recognize 92.18% pulmonary nodules,and the average detection time of each image is 0.035 s,which can effectively identify pulmonary nodules in CT images.At the same time,the auxiliary diagnosis system designed in this paper can realize the functions of patient information reading,lung parenchyma extraction,lung nodule detection and so on. |
PDF Full Text Request