Retinal Disease Classification

About the Dataset

The Retinal Fundus Multi-Disease Image Dataset (RFMiD) is a comprehensive dataset designed to advance research in the early detection and diagnosis of ocular diseases. According to the World Health Organization (WHO) World Report on Vision (2019), over 2.2 billion people worldwide are visually impaired, with at least 1 billion cases being preventable or unaddressed. This highlights the critical need for robust solutions in eye care.

The RFMiD dataset features 3,200 retinal fundus images captured using three different fundus cameras. These images have been expertly annotated for 46 unique ocular conditions, with consensus validation provided by two senior retinal specialists. This dataset serves as a valuable resource for developing machine learning models aimed at detecting and classifying multiple retinal pathologies, enabling researchers to combat preventable vision loss.

Key Features

1. Diverse Imaging: Fundus images captured using three distinct cameras, ensuring variety and robustness.
2. Expert Annotation: Adjudicated consensus by senior retinal experts ensures high-quality labelinga for 46 ocular conditions.
3. Comprehensive Scope: Supports both multi-disease classification and the differentiation between healthy and unhealthy retinas