Detection of diabetic retinopathy using image processing technique Test phase, srs, design phase and source code final deliverable
Domain/Category Image Processing
Introduction
Diabetic Retinopathy (DR) is an eye sickness caused by the increase of insulin in blood and ultimately prompts complete vision loss. It occurs when blood vessels in the retina swell and sometimes leak fluid or even close off completely. In different cases, irregular fresh vessels grow on the surface of the retina. In Pakistan, the number of patients with diabetes is high and they likewise have absence of information about the impact of diabetes on their retina and vision. Pakistan has around 7.1 million diabetic patients and very nearly 12000 patients experience the ill effects of extreme DR consistently. The ratio of ophthalmologists in hospitals of Pakistan is very low and many diabetic patients are unable to treat themselves lead to sudden vision loss. In rural areas this ratio is comparatively high in diabetic patients due to unavailability of well-established ophthalmological departments. Main objective of this project is to facilitate ophthalmologists is retinopathy process. System will able to tell user about current condition of patient retinal eye. System will able to perform computerized screening and diagnosis of anomalies (lesions) present in human retina through DR symptoms. The first one is Micro aneurysms (MAs), the earliest clinical sign of diabetic retinopathy; these occurs secondary to capillary wall out pouching due to pericyte loss. They appear as small red dot in the superficial retinal layer. The second one is Hemorrhage (H), they occur as MAs rupture in the deeper layer of the retina such as the inner nuclear and outer plexiform layers. The last is Exudates that are caused by the breakdown of blood retina barrier. It is a digital medicine project; student need to build an application to detect lesions by using any image processing technique or any artificial intelligence techniques.
Important note: for this project, student need to read at least 20 research papers and also select technique use to perform project functionalities.
Functional requirement
- They system shall verify the user with help of user database
- System Will allow user to add username and password.
- The system shall able to access patient database
- They system shall able to create a new folder so new patient data and retinal images can add through system window.
- User can access retinal image of required patient by using patient id through system.
- Complete user interface (with button and menus) must be functional.
- System will able to detect the abnormalities in blood vessels of retinal image. 8. System shall able to detect fluid leakage in retinal vessel 9. System shall able to detect swelling in retinal veins.
10.After detection of abnormalities in blood vessels system shall able to detect all lesions; Micro aneurysm, Hemorrhage, Exudates.
11.In result, system shall able to highlight abnormal vessels
12.System shall able to find present lesions in retinal image, in form of image.
13.Database of patients and users need to manage.
Data Collection
Data collection is most important task of this project. Correct data of retinal images is basic requirement if you want to get proper results. You may visit ophthalmological departments in hospitals for collection of data. Open-source data is also available but make sure it will be reliable.
Platform
Python, MATLAB or any other tool
Supervisor
Name: Anam Naveed
