A hybrid model for the detection of retinal disorders using artificial intelligence techniques

Area of Science:

• Ophthalmology
• Medical Imaging
• Artificial Intelligence

Background:

• Vision impairment prevalence is rising globally.
• Accurate and timely diagnosis of retinal disorders is crucial.
• Current diagnostic methods can be time-consuming and prone to human error.

Purpose of the Study:

• To develop an automated framework for classifying retinal disorders using optical coherence tomography (OCT) images.
• To compare the performance of various machine learning and deep learning models for retinal disorder classification.
• To improve the efficiency and accuracy of diagnosing conditions such as choroidal neovascularization, diabetic macular edema, and drusen.

Main Methods:

• A novel framework combining machine learning and deep learning was proposed.
• The InceptionV3 convolutional neural network was utilized as a feature extractor.
• Classifiers including Support Vector Machine (SVM), K-nearest neighbor (K-NN), Decision Tree (DT), and Ensemble Model (EM) were employed.
• A dataset of 18,000 OCT images was used for evaluation.

Main Results:

• The automated system achieved high classification accuracies for all tested models.
• Support Vector Machine (SVM) achieved 99.43% accuracy.
• K-nearest neighbor (K-NN) achieved 99.54% accuracy.
• Decision Tree (DT) achieved 97.98% accuracy.
• Ensemble Model (EM) achieved 99.31% accuracy.

Conclusions:

• The proposed methodology offers a promising approach for automated retinal disorder identification and classification.
• This automated system has the potential to significantly reduce human error in diagnosis.
• The framework can lead to substantial time savings in clinical practice, facilitating earlier treatment.