شماره راهنما :
1113 دكتري
پديد آورنده :
رشنو، عبدالرضا
عنوان :
ارائه يك روش خودكاردر فضاي نتروسافيك جهت بخش بندي ناهنجاري هاي فلويد درتصاوير ۳ بعدي OCT شبكيه
محل تحصيل :
اصفهان: دانشگاه صنعتي اصفهان، دانشكده برق و كامپيوتر
صفحه شمار :
پانزده، [۱۱۵]ص.: مصور، جدول، نمودار
استاد راهنما :
سعيد صدري، بهزاد نظري
توصيفگر ها :
شبكيه چشم , تصوير برداري OCT از شبكيه , بخش بندي فلويد , تئوري نتروسافيك , تئوري گراف , الگوريتم هاي كوتاهترين مسير گراف
استاد داور :
نادر كريمي، رسول امير فتاحي، عمادالدين فاطمي زاده
تاريخ ورود اطلاعات :
1396/10/24
رشته تحصيلي :
برق و كامپيوتر
دانشكده :
مهندسي برق و كامپيوتر
كد ايرانداك :
ID1113 دكتري
چكيده انگليسي :
Retinal 3D OCT segmentation of fluid abnormlities with an automated method in Neutrosophic domain Abdolreza Rashno a rashno@ec iut ac ir December 31 2017 Department of Electrical and Computer Engineering Isfahan University of Technology Isfahan 84156 83111 Iran Degree PhD Language Farsi Supervisors Prof Saeed Sadri and Dr Behzad Nazari Advisor Dr Hossein RabbaniAbstract Retina is one of eye components which consists different layers The main task of retina is to convert light energyinto neural signals interpretable for brain Age related Macular Degeneration AMD and Diabetic Macular Edema DME are the most common diseases in retina Imaging from layers and blood vessel structures is one of the most importantmethods used for diagnosis and treatment of retina diseases Optical Coherence Tomography OCT is the newest and themost advanced imaging method which creates high resolution images from tomographic sections and provides importantinformation related to layered structures of the retina The main objective of retina image processing is to create a retinaimage analysis system for manifesting retina diseases and its corresponding symptoms such as fluid cyst regions which areone of the main symptoms for AMD and DME diseases The main contribution of this thesis is fluid segmentation in AMDand DME subjects In this work a new method to transform images from pixel to neutrosophic domain and 2 automatedmethods for fluid segmentation are presented in neutrosophic domain For fluid segmentation in DME subjects first lastand 2 middle layers are segmented with graph shortest path methods based on new definitions of graph weights First andlast layers are used for removing background from images and middle layers are applied in post processing In the nextstep a new cost function is proposed and minimized which leads to image clustering Number of clusters is calculatedautomatically Finally regions between middle layers are removed For fluid segmentation in AMD subjects first and lastlayers are segmented and last layer is flattened Then seed points for fluid and tissue regions are initialized automaticallyfollowed by proposing a new cost function for graph cut in neutrosophic and kernel space Cost function minimizationleads to binary segmentation of images into fluid and tissue regions Important properties of the proposed steps are provedand quantitative performance of each step is analyzed separately Furtheremore clinical importance of fluid regions isinvestigated and the correlation of fluid regions with vision acuity are analyzed for diagnisis of retina diseases Key Words Retina OCT imaging fluid segmentation neutrosophic theory graph theory graphshortest path methods
استاد راهنما :
سعيد صدري، بهزاد نظري
استاد داور :
نادر كريمي، رسول امير فتاحي، عمادالدين فاطمي زاده
