Please use this identifier to cite or link to this item:
http://hdl.handle.net/10603/372503
Title: | Retinal Image Analysis for Screening of Diabetic Retinopathy |
Researcher: | Biyani Roopali Shyamsundar |
Guide(s): | Patre Balasaheb M. |
Keywords: | Engineering Engineering and Technology Instruments and Instrumentation |
University: | Swami Ramanand Teerth Marathwada University |
Completed Date: | 2021 |
Abstract: | Diabetes is one of the most rapidly growing global health threat of the 21st century. newlineDiabetic Retinopathy (DR) occurs as the direct outcome of uncontrolled diabetes newlinecausing microvascular damage to blood retinal barrier, leading to leakage and newlineblockage in retinal capillaries. This is major cause of visual impairment or loss newlineof vision in working-age population across the globe. DR perceive no symptoms newlinein prior, later progresses in stages and also depicted by several different lesions newlineoccurring on retina of the eye viz. microaneurysms (MA), haemorrhages (HE), newlinesoft exudates (SE) and hard exudates (EX). newlineEarly diagnosis and timely medical intervention of DR can prevent sight impairment newlineand blindness. Regular eye screening should therefore be an essential component newlineof routine diabetes care provided by primary healthcare professionals. newlineHowever, the limited availability of resources and healthcare personnel make the newlinemass screening difficult. There is a grave need for computer aided screening and newlinediagnosis for precise and timely treatment. The emerging computer vision and pattern newlinerecognition technologies provide opportunities to the biomedical researchers newlineand computer scientists to assist the healthcare professionals and ophthalmologists. newlineThe main objective of the work is to contribute to the computer aided screening newlineand diagnosis of various lesions in DR. The extraction of normal (optic disk (OD) newlineand retinal vessels) and abnormal (EX, MA, HE) structures from the retinal images newlineis the main aim of this research. newlineThe first contribution is an attempt to help the ophthalmologists in the DR screening newlineprocess to detect EX from non-mydriatic low-contrast retinal digital images newlinefaster and more easily. This contribution proposes a method using K-means clustering newlineand morphological image processing for detection of EX. In addition, OD newlineis detected and eliminated from the retinal images to avoid its intervention in EX newlinedetection. newlineThe screening of MA is an earliest measure to screen and treat DR. The second newlinecontribution focusses on auto |
Pagination: | 132p |
URI: | http://hdl.handle.net/10603/372503 |
Appears in Departments: | Department of Instrumentation Engineering |
Files in This Item:
File | Description | Size | Format | |
---|---|---|---|---|
01_title.pdf | Attached File | 269.5 kB | Adobe PDF | View/Open |
02_declaration.pdf | 46.31 kB | Adobe PDF | View/Open | |
03_certificate.pdf | 58.06 kB | Adobe PDF | View/Open | |
04_acknowledgements.pdf | 49.05 kB | Adobe PDF | View/Open | |
05_abstract.pdf | 72.37 kB | Adobe PDF | View/Open | |
06_contents.pdf | 108.18 kB | Adobe PDF | View/Open | |
07_list of figures.pdf | 151.17 kB | Adobe PDF | View/Open | |
08_list of tables.pdf | 131.86 kB | Adobe PDF | View/Open | |
09_list of algorithms.pdf | 45.91 kB | Adobe PDF | View/Open | |
10_abbreviations.pdf | 64.96 kB | Adobe PDF | View/Open | |
11_chapter 1.pdf | 2.68 MB | Adobe PDF | View/Open | |
12_chapter 2.pdf | 127.8 kB | Adobe PDF | View/Open | |
13_chapter 3.pdf | 14.32 MB | Adobe PDF | View/Open | |
14_chapter 4.pdf | 4.77 MB | Adobe PDF | View/Open | |
15_chapter 5.pdf | 3.93 MB | Adobe PDF | View/Open | |
16_chapter 6.pdf | 13.22 MB | Adobe PDF | View/Open | |
17_cconclusion.pdf | 94.02 kB | Adobe PDF | View/Open | |
18_summary.pdf | 94.02 kB | Adobe PDF | View/Open | |
19_bibliography.pdf | 108.63 kB | Adobe PDF | View/Open | |
80_recommendation.pdf | 393.41 kB | 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: