Automated Diagnosis of Diabetic Retinopathy using FPGA

Diabetic Retinopathy (DR) has always been a common disease in people suffering from diabetes. Diagnosing it in early stages is very essential to avoid a diabetic patient from going completely blind. Several techniques have been developed for diagnosing DR. Previously, the images of fundus camera wer

Project Title

Automated Diagnosis of Diabetic Retinopathy using FPGA

Project Area of Specialization

Artificial Intelligence

Project Summary

Diabetic Retinopathy (DR) has always been a common disease in people suffering from diabetes. Diagnosing it in early stages is very essential to avoid a diabetic patient from going completely blind. Several techniques have been developed for diagnosing DR. Previously, the images of fundus camera were processed for the diagnosis of DR manually and it used to take an extended period of time and it is prone to human error. It calls for the need of developing a system that is automated, highly accurate and fast. For achieving highly accurate results, sophisticated image processing algorithms are incorporated. These algorithms are very complex and require high computational time. To reduce the time and power consumption in computation of these algorithms, we will implement the system on FPGA. FPGA is considered because of its inherent parallelism for flexibly implementing image processing with high performance and less power consumption. Also they allow our systems and devices to be portable so that they can be used at Point of Care (PoC) such as remote areas that lack proper health care facilities.

Project Objectives

1. Develop a highly accurate system for Diabetic Retinopathy diagnosis.

2. Develop a system that can run in real time and provides high speed solutions with low power consumption.

3. Develop an indigenous system to facilitate doctors and patients in Pakistan for low cost automated non-invasive diagnosis.

Project Implementation Method

1.IMAGE ACQUISITION: Take an image, obtained by the fundus camera of size 600x700x3.

2.PREPROCESSING (to improve image quality): Create an average level of fundus image background, Isolate the green channel, finally take the negative of resultant image.

3. VESSEL SEGMENTATION: After preprocessing, vessels are segmented through morphological processing.

4.OPTIC DISC DETECTION:After preprocessing, in parallel of vessel segmentation, optic disc detected through bright region detection. 

5. POST PROCESSING: Overlap the results of vessel segmentation and optic disc detection.

6. IMAGE INTERPRETATION: Create a for loop and give the following conditions: if(1) added in white, if (0) added in black. Then generate the ratios, and finally on the basis of result decide either DR present or not.

Benefits of the Project

1. A highly accurate system for Diabetic Retinopathy diagnosis.

2. This is a system that can run in real time and provides high speed solutions with low power consumption.

3. This is an indigenous system to facilitate doctors and patients in Pakistan for low cost automated non-invasive diagnosis.

Technical Details of Final Deliverable

The final deliverable of our final year project would be a a method to identify diseased fundus images and distinguish them from normal ones, and implementing this method on an FPGA KIT using xylinx to reduce time consumption of this process.

Final Deliverable of the Project

Hardware System

Type of Industry

Medical

Technologies

Others

Sustainable Development Goals

Good Health and Well-Being for People

Required Resources

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