Research on Computer Aided Diagnosis for Cerebral Aneurysm Based on DSA Images
|Course||Applied Computer Technology|
|Keywords||Computer-aided diagnosis Cerebral aneurysms Suspected lesions positioning Feature extraction Classification|
With the successful application of the rapid development of computer technology and medical imaging in clinical diagnosis , computer-aided diagnosis technology is growing . Cerebral aneurysms in clinical practice or through the doctor observed angiograms diagnosis , the drawback is the poor accuracy and objectivity . Cerebral aneurysms computer-aided diagnosis technology is still in the research stage , there is no clinical application . This paper summarizes the techniques and methods of computer-aided diagnosis system is widely used in clinical and proposed a set of auxiliary diagnosis of cerebral aneurysms program , including cerebral aneurysms suspected lesions positioning and based on the statistical classification of the features of cerebral aneurysms . Firstly, the basics of cerebral aneurysms and DSA imaging principle , doctors diagnosed cerebral aneurysms and methods . On this basis , this paper proposes suspected lesion localization method based on the characteristics of a skeleton , the method can locate the location of the suspected lesion area , providing important details to the doctor DSA image , its meaning is to avoid the image detail is ignored doctors reasons of their own , thus assist doctors diagnosis of cerebral aneurysms . Same time, analysis of the the DSA brain blood vessel image characteristic and the binding direction of the memory , the on fast boundary tracking cerebral blood vessel image , the information of the the cerebrovascular boundary geometry , shape feature extraction of cerebral aneurysms . This paper presents a cerebral aneurysms recognition method based on the statistical classification algorithm , this method uses the RBF neural network , SVM support vector machine algorithm and Bayesian algorithm to classify the shape feature of cerebral aneurysms , and analysis and comparison of the three categories the results of the algorithm . The method provides doctors cerebral aneurysms auxiliary diagnostic information to reach the auxiliary diagnostic purposes . Finally , the work of the graduation subject is outlined further research ideas .