AI for Health Healthon 2026 October 2026

Ibsar Al-Ayn إبصار العين

AI for early detection of diabetic retinopathy — a portable, smartphone-based retinal camera integrated with an explainable AI platform that brings screening closer to patients.

90%Accuracy on held-out test images
4Deep-learning models in the ensemble
5ICDR-based severity levels
AR/ENBilingual clinical reports
The Story

Innovation beyond winning

Not every success is measured by an award. Sometimes, success is identifying a real societal challenge, developing a meaningful solution, and taking the first step toward making an impact.

We are proud to share Ibsar Al-Ayn (إبصار العين), an AI-enabled portable screening solution for the early detection of diabetic retinopathy, developed by our PhD student Muhammad Muzzammil Auzine under the supervision of Prof. Tanzila Saba at AIDA Lab, Prince Sultan University.

The challenge

Diabetic retinopathy remains a major cause of preventable blindness, and Saudi Arabia has a significant need for accessible and timely diabetic eye screening. Ibsar Al-Ayn aims to bring screening closer to patients through a portable, smartphone-based retinal camera integrated with an explainable AI platform.

Presented at Healthon 2026

The team achieved 90% accuracy on held-out test images and presented a complete solution at Healthon 2026, held 1–4 October 2026 at University Medical City, King Saud University. Although the team did not win the competition, this journey is a success. Competitions may end with awards, but innovation begins when an idea has the potential to move beyond the competition and serve society.

  • Working web application
  • Trained AI models
  • Complete 3D design of the capture device
  • Demonstration video
  • Business model
  • A main milestone toward real-world impact

Why it matters

There is still enormous scope for AI in diabetic healthcare in Saudi Arabia — from early screening and prevention to personalized care and clinical decision support. Projects like Ibsar Al-Ayn demonstrate how our students can contribute to this rapidly evolving space.

The Solution

Two parts, one kit

A low-cost capture device in the clinic, connected to an explainable AI platform that grades, reports and refers.

1

Handheld retinal camera

Smartphone opticsRed previewWhite flashBluetooth trigger
2

AI web platform

Quality gate4-model ensembleGrad-CAMReportsReferrals
Key Features

From capture to referral

Image quality assessment

Automated retinal image quality checks before any grading takes place.

Five-level DR grading

Diabetic retinopathy severity graded on the international ICDR scale.

Four-model ensemble

An ensemble of four deep-learning models for more robust predictions.

Explainable AI

Grad-CAM heatmaps show clinicians where the model is looking.

Arabic & English reporting

Bilingual outputs designed for patients and clinicians in Saudi Arabia.

PDF reports & referrals

Automated PDF reports with referral tracking to specialists.

What's Next

Not another competition — real-world validation

Build the physical prototype

Turn the complete 3D design into a working handheld capture device.

Validate on Saudi data

Validate the models on Saudi retinal datasets.

Pilot in primary care

Ultimately pilot the solution in a primary-care setting.

We are thankful to the Research and Initiatives Center (RIC), Prince Sultan University for bringing this vision forward.

This is what we want from innovation at AIDA Lab

Identify a real problem Develop an intelligent solution Validate it Take it toward society