Artificial intelligence-based decision-making for age-related macular degeneration

Area of Science:

• Ophthalmology
• Medical Imaging
• Artificial Intelligence

Background:

• Artificial intelligence (AI), particularly convolutional neural networks (CNNs), shows promise for improving healthcare quality and medical workflows.
• Cloud-based AI tools are of significant interest for telemedicine applications, enabling remote medical care via electronic interfaces.

Purpose of the Study:

• To develop and evaluate an AI- and cloud-based telemedicine tool for the diagnosis and treatment recommendation of age-related macular degeneration (AMD).
• To assess the diagnostic performance of the AI system compared to medical professionals.

Main Methods:

• Trained three types of CNNs using a dataset of 35,900 labeled optical coherence tomography (OCT) images from AMD patients.
• Developed a cloud-based platform for telemedicine interaction, allowing patients to upload OCT images for AMD verification.

Main Results:

• The AI system achieved a high detection accuracy (over 90%), comparable to retinal specialists (92.73% and 91.90%) and significantly superior to medical students (69.4% and 68.9%).
• The AI platform provided treatment recommendations on par with those of retinal specialists.
• A website was developed for cloud-based AI analysis of OCT images for AMD diagnosis.

Conclusions:

• An AI-based cloud service offers a practical solution for medical imaging diagnostics and telemedicine.
• The developed AI platform can accurately diagnose AMD and suggest appropriate treatments, improving accessibility to eye care.
• The website provides a direct channel for patients to utilize AI for AMD screening and management.