Quantitative Analysis And Evaluation System Of Stroke Based On Images

Stroke has been the first cause of death in our residents already.Many patients have very serious consequences because of missing the best treatment time.At present,the diagnosis of clinicians depends on medical imaging.Therefore,it is meaningful to design a rapid,stable,quantitative and accurate imaging-based stroke quantitative analysis system.Not only can improve the efficiency of diagnosis and accuracy of the doctor,but also gain more valuable time for the treatment of patients to reduce the death of stroke,reduce postoperative impact on patients.Based on the medical image sequence and image post-processing technology,a quantitative analysis system of stroke based on multi-modality/multi-sequence images is proposed,and a software product is designed,sequence includes CT,T2WI,SWI,MRA and so on.This system has designed a series of functions for multi-sequence processing,such as multi-sequence comparison,positioning and synchronization according to the image registration technology,to meet the needs of doctors in dealing with multiple sequence images of stroke.At the same time,in the ease of use and user experience to do a detailed thinking,different from the previous medical imaging software.This paper mainly completed the following work:(1)The design and development of image-based stroke quantitative analysis system were completed.According to the needs of clinical imaging diagnosis of stroke,data management,image browsing and quantitative analysis were designed for the diagnosis of multi-mode/multi-sequence images.The system uses QT platform,C ??,Visual Studio2013 compiler,MySQL management database,DICOM medical image format,integrated ITK,VTK and other development kits.(2)Using image registration algorithms,design multi-modality/multi-sequence image contrast and positioning mode.According to the clinical requirements,two registration algorithms are provided to realize the 3D registration of multi-mode/multi-sequence images,based on the DICOM header file registration,and the mutual information optimization image registration method.Design two application modes,namely,image contrast mode and image positioning mode.(3)Using image segmentation algorithms to achieve brain tissue and stroke segmentation.In this paper,I used fuzzy mean clustering algorithm to realize the segmentation of stroke.One key to remove the skull in the sequence,having segmented out the gray matter,white matter,cerebrospinal fluid and other brain tissue.Finally,a quantitative analysis module of stroke was put forward.Three quantitative analysis modes,single sequence single analysis,single sequence multiple analysis and single sequence multiple analysis were designed.