DR Sense

Diabetic retinopathy is a diabetes related complication that affects eye. It causes damage to blood vessels of the light-sensitive tissue at the back of the eye (retina). At first diabetic retinopathy may cause no symptoms or mild vision problem but eventually it causes blindness.

Project Title

DR Sense

Project Area of Specialization

Artificial Intelligence

Project Summary

Diabetic retinopathy is a diabetes related complication that affects eye. It causes damage to blood vessels of the light-sensitive tissue at the back of the eye (retina). At first diabetic retinopathy may cause no symptoms or mild vision problem but eventually it causes blindness.

The patients suffering from prolong diabetes are the possible victims of diabetic retinopathy.  due to lack of early diagnosis and less control on patient’s glucose levels it becomes  one of the leading cause of  vision loss around the world. However, if diagnosed and treated early, it can prevent the further damage. Especially, populous and resource limited countries like Pakistan necessitates the development of more reliable and easy to access solution to detect DR.

In this regard, we are designing a mobile phone application in which a 30D-lense and the frame is attached to the andoid device in order to capture the fundus image of patient's retina and than the patient fundus image will upload and routed to server for processing. Using image processing, deep learning and AI techniques severity level of the disease will be identified and results will sent back to application. Once results are there it notifies the patient either he/ she is suffering from Low NPDR, Mild NPDR or Severe NPDR and  make suggestions to either visit nearby eye-care facility or to take precautionary measures.

Early detection can prevent the patients suffering or at risk of DR who are running out of resources like time and lack of accessibility to state of art technology to detect DR. The proposed solution provides improved accessibility and increased reliability, adding in both saving the time and underlying cost.

Project Objectives

The primary objective this project is:

• A way to device new plans for disease prevention
• An awareness about the diabetes related complications
• Increased Patients Satisfaction
• A way to manage overpopulated hospitals

Project Implementation Method

The project is divided into various phases each phase having its own tools and area of expertise.

1. The first phase includes understanding the different aspects of diabetic retinopathy which also includes understanding the data required for processing, for that a publically available dataset of around 10,000 fundus images is available
2. In second phase an algorithim for processing the images(image processing techniques)will be devised for the leasion detection of the disease from fundus captured images. MATLAB Tool will be used for image processing.
3. In the next phase classification will be performed on the extracted leasion and extracted feature numbered information of those leasions. By applying varius techniques(Neural Network) for that and the various tools are available such as tensorflow,MATLAB.
4. In this phase a mobile interface will be designed and implemented which provides ease of access to the user and can diganose the disease in fast reliable manner. Required tool for that is Android Studio IDE.

Benefits of the Project

Following are the benifits of DR Sense Project

1. An AI based technique will grade the  fundus  images based on its severity level of disease. without the involvement of a doctor or expert (rapid diagnosis) saving the patient's from overhead of waiting for doctors by automating all the procedural formalities such as appointment with doctor e.t.c.   
2. AN 30D-lense and associated frame is cost effective for capturing the fundus images of patient and this automated system can support the doctors in fast diagnosis and reliable diagnosis of diabetic retinopathy.
3. Its ability to prioritize the proliferative over non-proliferative patients mitigates the risk of
   blindness due to lack of early detection

Technical Details of Final Deliverable

The final deliverable of this project will be in the product shape which contains the assembeled 30D-Lense attached with the frmae and the android application. The dataset, patient fudus caputured images will be uplaod to the server. All processing will be performed on server side. An android application will used as interface to interact with user and the server.

Final Deliverable of the Project

HW/SW integrated system

Type of Industry

IT , Medical

Technologies

Artificial Intelligence(AI), Big Data

Sustainable Development Goals

Good Health and Well-Being for People

Required Resources

If you need this project, please contact me on contact@adikhanofficial.com