Development and Clinical Diagnostics of Ophthalmic Imaging Modalities based on Deep Learning System
According to World Health Organization (WHO) at present, at least 2.2 billion people around the world have vision impairments, of whom at least 1 billion have a vision impairment that could have been prevented. Traditionally the device used by all ophthalmologists for the diagnosis of eye diseases i
2025-06-28 16:32:02 - Adil Khan
Development and Clinical Diagnostics of Ophthalmic Imaging Modalities based on Deep Learning System
Project Area of Specialization Artificial IntelligenceProject SummaryAccording to World Health Organization (WHO) at present, at least 2.2 billion people around the world have vision impairments, of whom at least 1 billion have a vision impairment that could have been prevented. Traditionally the device used by all ophthalmologists for the diagnosis of eye diseases is Slit-lamp, which is un-economical. A better solution would be a low-cost device that can automatically detect the diseases and reducing the ophthalmologist’s workload and prevents vision damage of patients by suggesting a suitable cure for the patient.
The increasing prevalence of diabetes is leading to a rise in eye diseases, augmenting the risk of sight-threatening complications. The diabetic patients require checkup of eyes periodically, but this practice is not followed by patients, because the majority of them belongs to low and middle-class. As a consequence, the diseases were not diagnosed at an initial level that’s why patients will be at the risk of sight-threatening complication or vision loss.
 |
Our Idea is to develop a device that takes an image of the retina of the patient through a 20 D lens, LED light, and HD camera set-up. The image captured by the camera is processed in Raspberry Pi 4 through deep learning, if there would be any disease in the eye, it would be detected by the Raspberry Pi using a trained model and data set and displayed at the LCD (HMI) with the recommended cure. This device will be very helpful for rapid and automatic detection of critical diseases and assists in reducing the ophthalmologist’s workload and prevents sight-threatening complications.
The purpose of this project is to build a low-cost device that can detect various eye diseases correctly and recommend an effective cure by means of image processing using Deep Learning (DL). The device will make a prominent difference in the field of eye health care.
The goal of this project is to introduce deep learning in the eye health care system by using a Convolutional Neural Network Algorithm in order to build-up a fast, low-cost, and productive device.
The existing eye health care device (Slit Lamp) has some limitations like being expensive, fails to detect the correct eye diseases, it can only provide a clear image of the retina. This project would address these limitations through deep learning.
Project Implementation MethodWhen it comes to the implementation of the proposed Deep Learning-based eye disease diagnosis device, there would be many challenging tasks because we plan to introduce deep learning for the correct diagnosis of the patient’s eye.
In the past two years, a number of Deep Learning models with impressive performance have been developed for automated detection. Newly developed deep learning models have high efficiency and sensitivity.
Our Idea is to make a device that takes an image of the retina of the patient through a 20 D lens, LED light, and HD camera set-up. The image captured by the camera is processed through deep learning, if there would be any disease in the eye, it would be detected and displayed at the LCD with the recommended cure. This device will be helpful for rapid and automatic detection of critical diseases and assists in reducing the ophthalmologist’s workload and prevents vision damage of patients
The brain of our device is the Raspberry Pi 4 which carries out all the necessary action. Firstly, the data analytics would be done with data set by using EDA (Exploratory Data Analysis). The data would be then conveyed to the Cloud for fast processing, model creation, and training of the model. The Algorithm, which would be used in the deep learning model is “Convolutional Neural Network (CNN)”.
Benefits of the ProjectAccording to WHO, about 422 million people worldwide have diabetes. The increasing prevalence of diabetes is leading to a rise in eye diseases, augmenting the risk of sight-threatening complications. According to the survey in 2017, diabetic retinopathy was reported by 40.9% of patients with type 1 diabetes and 9.8% of patients with type 2 diabetes, 35.8% and 12.6% of all participants reported cataract and glaucoma, respectively.
The diabetic patients need checkup of eyes periodically, but this practice is not followed by patients, because the majority of them belongs to low and middle-class. As a consequence, the diseases were not diagnosed at an initial level which’s why patients will be at the risk of vision loss.
According to the survey, 90% of the patients with diabetes can avoid eye disease development through early detection. So, we come up with a device that is very cost-effective as compared to slit-lamp, can diagnose eye diseases, and recommend a suitable cure.
Technical Details of Final DeliverableThe final deliverables would consist of a Hardware system along with the trained model of the system. The hardware system would consist of a pre-programmed Raspberry Pi 4, 20D Bio Lens, LED Light HD Camera setup along with LCD.
We would deliver the trained model of the system on which the system would make correct eye disease detection.
Final Deliverable of the Project Hardware SystemCore Industry HealthOther Industries Medical Core Technology Artificial Intelligence(AI)Other TechnologiesSustainable Development Goals Good Health and Well-Being for PeopleRequired Resources| Elapsed time in (days or weeks or month or quarter) since start of the project | Milestone | Deliverable |
|---|---|---|
| Month 1 | Mapping of Project and procurement of equipment | Finalized project flow chart and procurement of equipment |
| Month 2 | Initiation of Deep Learning Skills | Beginner level Project Oriented Deep Learning Skills |
| Month 3 | Advancement of Deep Learning Skills | Mastering the Project Oriented Deep Learning Skills |
| Month 4 | Data Collection | Collection Of Wide Range Of Data |
| Month 5 | Data Analysis | Data Analytics |
| Month 6 | Deep Learning Model Creation | Development & Training of Deep Learning Model |
| Month 7 | Pondering Of Hardware | Determination of suitable Hardware |
| Month 8 | Development of Hardware (Initial stage) | Building-up of Hardware |
| Month 9 | Development of Hardware (Final stage) | Building-up of Hardware |
| Month 10 | Monitoring & Implementation | Monitoring, Implementation & Closure Of Project |