Automated Diagnosis of Retinal Diseases using Optical Coherence Tomography Images and Volumes using Deep Learning

Show simple item record

dc.contributor.author Das, Vineeta
dc.date.accessioned 2022-05-18T07:18:33Z
dc.date.available 2022-05-18T07:18:33Z
dc.date.issued 2021
dc.identifier.other ROLL NO.166102012
dc.identifier.uri http://gyan.iitg.ac.in/handle/123456789/1988
dc.description Supervisors: Samarendra Dandapat and Prabin K Bora en_US
dc.description.abstract This dissertation focuses on developing efficient and reliable classification methods for the automated diagnosis of retinal diseases using optical coherence tomography (OCT) images and volumes. The proposed methods take into consideration the pathological evidence of the retinal diseases and handle the imaging bottlenecks (speckle noise and poor resolution) of OCT to develop robust deep learning (DL) based classifiers. Specifically, the thesis discusses the usefulness of the multi-scale feature learning strategy to efficiently encode the diverse pathological characteristics of the retinal lesions for accurate classification. An unsupervised simultaneous denoising and super-resolution method is designed to improve the visualization of the minute clinical manifestations that occur at early stages of the retinal diseases for their efficient detection. In the end, an OCT volume classification method is proposed that automatically identifies the salient pathological images of the volume and provides them more importance during the diagnosis process. The developed methods have been evaluated on large-scale publicly available databases. The proposed approaches outperform the existing methods in terms of diagnostic accuracy and computational complexity making them suitable for reliable and fast diagnosis in eye clinics and hospitals. en_US
dc.language.iso en en_US
dc.relation.ispartofseries TH-2540;
dc.subject ELECTRONICS AND ELECTRICAL ENGINEERING en_US
dc.title Automated Diagnosis of Retinal Diseases using Optical Coherence Tomography Images and Volumes using Deep Learning en_US
dc.type Thesis en_US


Files in this item

This item appears in the following Collection(s)

Show simple item record

Search


Browse

My Account