Please use this identifier to cite or link to this item:
http://hdl.handle.net/10603/291003
Title: | Anatomical structure segmentation in retinal images with some applications in disease detection |
Researcher: | Chakravarty Arunava |
Guide(s): | Sivaswamy Jayanthi |
Keywords: | Computer Science Engineering and Technology Imaging Science and Photographic Technology |
University: | International Institute of Information Technology, Hyderabad |
Completed Date: | 2020 |
Abstract: | Color Fundus (CF) imaging and Optical Coherence Tomography (OCT) are widely used by ophthalmologists to visualize and quantify the structural deformations that characterize different retinal diseases. In this thesis, we propose different frameworks for the automatic extraction of boundaries of relevant anatomical structures in CF and OCT images with potential applications in the detection of retinal diseases such as Glaucoma, Diabetic Macular Edema (DME) and Age-related Macular Degeneration (AMD). newlineFirst, we address the problem of the segmentation of Optic Disc (OD) and Optic Cup (OC) in CF images to aid in the detection of Glaucoma. A novel boundary-based Conditional Random Field (CRF) framework is proposed to jointly extract both the OD and OC boundaries by modelling the drop in depth between them. newlineNext, we consider the task of the intra-retinal tissue layer segmentation in OCT images which is essential to quantify the structural changes caused by AMD and DME. We propose a supervised CRF framework to jointly extract the eight layer boundaries. The appearance features for each layer and the relative weights of the shape priors are learned using a Structured Support Vector Machine formulation. newlineNext, we propose a novel deep Recurrent Neural Network architecture called the Recurrent Active Contour Evolution Network (RACE-net) which models the level set based deformable models for medical image segmentation. It demonstrated a consistent performance across a diverse set of segmentation tasks viz. the extraction of OD and OC in CF images, cell nuclei in histopathological images and left atrium in cardiac MRI volumes. newlineWe close this dissertation with some illustrative applications in the detection of retinal diseases. We explore and benchmark two classification strategies for the detection of glaucoma from CF images based on deep learning and handcrafted features respectively. We also construct a Normative Atlas for the macular OCT volumes to aid in the detection of AMD. newline newline |
Pagination: | |
URI: | http://hdl.handle.net/10603/291003 |
Appears in Departments: | Computer Science and Engineering |
Files in This Item:
File | Description | Size | Format | |
---|---|---|---|---|
01_title.pdf | Attached File | 50.83 kB | Adobe PDF | View/Open |
02_certificate.pdf | 30.42 kB | Adobe PDF | View/Open | |
03_acknowledgments.pdf | 91.27 kB | Adobe PDF | View/Open | |
06_chapter1.pdf | 2.37 MB | Adobe PDF | View/Open | |
07_chapter2.pdf | 4.95 MB | Adobe PDF | View/Open | |
08_chapter3.pdf | 7.73 MB | Adobe PDF | View/Open | |
09_chapter4.pdf | 5.98 MB | Adobe PDF | View/Open | |
10_chapter5.pdf | 5.52 MB | Adobe PDF | View/Open | |
11_chapter6.pdf | 448.19 kB | Adobe PDF | View/Open | |
80_recommendation.pdf | 3.98 MB | Adobe PDF | View/Open |
Items in Shodhganga are licensed under Creative Commons Licence Attribution-NonCommercial-ShareAlike 4.0 International (CC BY-NC-SA 4.0).
Altmetric Badge: