Technology Overview

The Artificial Intelligence (AI) powered eye imaging diagnostic solution uses deep learning and machine learning models to grade eye fundus images. It provides results that identify the presence and the severity of different diseases. These AI diagnostic solutions are embedded in a platform with modern and user-friendly web app and mobile app interfaces to present a low-cost solution for the users within the eye health industry.

The web app offers a comprehensive patient management system for general practitioners (GPs), ophthalmologists and optometrists. In addition to the AI diagnostic solutions, other functionalities include a statistical dashboard, patient health data management and teleconference (for further eye-grading and consultations).

The mobile app, a lightweight version of the web app, connects patients with the healthcare professionals listed above. It also provides patients access to their eye exam results and reports which enables them to analyse and understand their health with ease.

A bridging technology is also available which automates the image uploading process between fundus cameras and the platform.

TeleMedC is open to collaborating with fundus camera providers, research scientists, labs, optical shops, pharmacists and more.

Technology Features, Specifications and Advantages

• User-friendly interface that facilitates a seamless user journey and hides the complexity of the embedded AI diagnostic solutions
• Quick drag and drop to upload fundus images
• Results of the presence and severity of disease return within seconds of uploading.
• A comprehensive report, containing results and educational material, is automatically generated

The AI diagnostic solutions currently target multiple diseases including:

Diabetic Retinopathy (DR): One of the models instantly generates a YES/NO result for DR and the other checks for DR severity, whether it’s mild, moderate, severe NPDR or worse. Our DR model is one of the most accurate image diagnostic AI services in the world with an accuracy of 97% sensitivity and 92% specificity.

Glaucoma: The model provides automatic optic disc and cup segmentation results by calculating the cup-disc ratio (CDR) and fellow CDR asymmetry for glaucoma screening.

Age-related Macular Degeneration (AMD): The model provides a YES/NO solution to end users’ AMD condition.

These AI models have regulatory certification approved by: ISO 13485, Therapeutic Goods Administration (TGA) Australia, Health Sciences Authority (HSA) Singapore and CE Germany.

Potential Applications

With the right algorithms in place, these technologies can be applied to detect the early stages of any other disease. The potential of these products is huge, for instance:

• Referring urgent patients via scheduling and teleconferencing
• Turning expensive medical hardware into cheap and affordable IOP device
• Patient self-management platform
• Education and nutritional information

Customer Benefits

80% of diabetic patients will eventually develop diabetic retinopathy and that may lead to blindness. The goal is set to provide low-cost diagnostic solutions for point-of-care screening and virtual management of eye and chronic disease worldwide. By deploying fundus cameras in each region, clinics, optometrists and other users within the eye industry can share the same camera, hence reaching out to more patients and being extremely cost-effective.