Identification of cataract and post-cataract surgery optical images using artificial intelligence techniques

Area of Science:

• Ophthalmology and Artificial Intelligence
• Medical Image Analysis
• Computer Vision in Healthcare

Background:

• Cataract, a leading cause of blindness in the elderly, involves gradual clouding of the eye's lens.
• Early detection and monitoring of cataract progression and surgical success are crucial for preserving vision.
• Existing diagnostic methods may require further enhancement for accuracy and efficiency.

Purpose of the Study:

• To develop and evaluate an artificial intelligence (AI) system for detecting cataracts using optical images.
• To assess the system's capability in evaluating the efficacy of post-cataract surgery.
• To classify optical images into normal, cataract, and post-cataract surgery categories.

Main Methods:

• Image processing techniques applied to raw optical images.
• Fuzzy K-means clustering algorithm used for feature detection.
• Backpropagation algorithm (BPA) employed for classification using an Artificial Neural Network (ANN) classifier.
• Dataset comprised 140 optical images across three classes.

Main Results:

• The ANN classifier achieved an average detection rate of 93.3% for normal, cataract, and post-cataract images.
• The proposed system demonstrated high performance with 98% sensitivity and 100% specificity.
• Results indicate clinical significance and potential for real-world application.

Conclusions:

• The developed AI system effectively detects cataracts and assesses post-surgical outcomes from optical images.
• The system's high sensitivity and specificity suggest its clinical utility in ophthalmology.
• This AI-driven approach offers a valuable tool for monitoring eye health and surgical interventions.