Cad Aspects in Acute Ischemic Stroke Patient Evaluation

In recent years, stroke mortality rate although there is a downward trend, but probability of occurrence still growing up. Due to part of interpretation in clinical practice, the relevant personnel to define the area and scope of stroke are not consistent. Therefore this study, focus on the cerebral arteries and posterior cerebral arteries of the brain, seriousness level of ischemic stroke, the same of evaluation standard in clinical practice with the brain boundaries and part of ischemia. However, this study will develop a computer-aided detection system for ischemic stroke diagnosis, also able to effectively determine the seriousness level of ischemic, giving effective dose of rt-PA treatment in limit time. This project will build the system through image processing technology. According a preliminary study, first of all, an adaptive noise filter will be used in the non-enhanced CT image to filtering noise, and bi-level and regional growth methods used to obtain effective image information, and then the ischemic contour and area can be obtained from the texture parameters and automatic selection. Finally, to calculate ASPECTS score by compared to the left brain and right side and the support the vector machine used as a follow-up parameters of the statistical analysis. The CT and MRI images of 600 sets (each set including 8 section images and exclude images with tumor, edema and hemorrhage) were expected to be used in the project. 20 sets CT and MRI were used as the training group and 580 sets as test sets. Doctors make naked eye interpretation ASPECTS score and professional knowledge to determine the location and verify the results of the system. DICOM image reader, parameter adjustment, disease identification and identification of suspicious areas will be included in the platform design for this system. We expect this computer-aided detection system can help physicians to more quickly and accurately determine the pathology, and provide objective information to less experienced physicians for diagnosis.