Strategic Insights

Sector: Healthcare
Service: AI & Emerging Technology Integration
Location: Riyadh, Saudi Arabia

Challenge:
A prominent Riyadh-based hospital faced challenges in accurately and swiftly diagnosing eye conditions such as glaucoma and diabetic retinopathy, relying heavily on manual assessments that were time-consuming and prone to human error.

Solution:
AyadElm implemented an AI-driven diagnostic tool utilizing a U-Net architecture for medical image segmentation. This solution enabled automated identification and segmentation of retinal abnormalities in high-resolution images.

Results:

• F1 Score: 0.837

• Intersection over Union (IoU): Above 80%

• Diagnosis Time per Image: Approximately 25 milliseconds

• Impact: Enhanced diagnostic accuracy, reduced clinician workload, and expedited patient care.