Detection of Diabetic Macular Edema in Optical Coherence Tomography Image Using an Improved Level Set Algorithm

Area of Science:

• Ophthalmology
• Medical Imaging
• Computer Vision

Background:

• Diabetic macular edema (DME) is a leading cause of vision loss in diabetic retinopathy patients.
• Current DME detection in Optical Coherence Tomography (OCT) images relies on time-consuming manual analysis by clinicians.
• Automated detection is crucial for efficient, large-scale screening and timely intervention.

Purpose of the Study:

• To develop and evaluate a novel algorithm for automated DME detection and segmentation in OCT images.
• To compare the proposed algorithm's performance against existing level set methods and manual segmentation.
• To assess the algorithm's potential for mass screening of diabetic retinopathy.

Main Methods:

• Proposed the SBGFRLS-OCT algorithm, combining K-means clustering and an improved Selective Binary and Gaussian Filtering regularized level set (SBGFRLS) approach.
• Compared SBGFRLS-OCT with Chan-Vese (C-V), geodesic active contour (GAC), and the original SBGFRLS algorithms.
• Evaluated precision, sensitivity, and specificity against manual segmentation by clinicians.

Main Results:

• SBGFRLS-OCT demonstrated improved accuracy and reduced processing time compared to C-V, GAC, and SBGFRLS algorithms.
• The algorithm achieved high performance metrics: 97.7% precision, 91.8% sensitivity, and 99.2% specificity.
• Results indicate performance comparable to manual segmentation by experienced clinicians.

Conclusions:

• The novel SBGFRLS-OCT algorithm provides an accurate and efficient method for DME detection in OCT images.
• This automated approach can significantly aid in the early diagnosis of diabetic retinopathy and prevention of blindness.
• The algorithm shows promise for facilitating mass screening programs for diabetic retinopathy.