A Level Set Method For The Gastrointestinal Tumors Detection

The gastrointestinal tumors can cause more other diseases in modern society.To solve the problem,wireless capsule endoscopy(WCE)has been wildly used to detect tumors in the digestive tract and it exhibits as an effective detecting means.However a considerable challenge has presented for a doctor since the WCE takes about 60,000 pictures for each person.Recently,there are barely researching reports in gastrointestinal tumors images(GTI).In this thesis,we propose a GTI processing which can reduce the doctors' workload.Firstly,based on the characteristics of WCE image and Chan-Vese model,we use multiphase level set method to get the interesting region.Therefore,we obtain two matrixes,in which one contains the whole cell's boundary,and the other contains the boundary of some cells.Secondly,with respect to the cells' boundary,it is necessary to do further processing,which ensures the boundary of some cells is a discrete region.Mathematical Morphology brings a fancy result during the discrete processing.As for a tumor,two features are accepted in the domain,one of them is that the color of the tumor is light,and the other the shape is similar to an ellipse.At last,we consider every discrete region according to the tumor's features to judge whether a tumor appears in the image or not.Our method has a desirable performance in the presence of common GTI.For some non-convex GTI,we utilized a traditional way(SVM and LBP)as a second processing,which increased the coverage and accuracy.Experiments show that our method has a high coverage without any learning-based classifiers,which saves a lot time and reduces a lot workload.Therefore,the proposed method has a good practical application for assisting physicians in detecting tumors in the GTI image.