Stochastic Differential Equations for Automatic Quality Control of Retinal Optical Coherence Tomography images

Area of Science:

• Ophthalmology
• Medical Imaging
• Biomedical Engineering

Background:

• Optical coherence tomography (OCT) provides high-resolution retinal images crucial for diagnostics.
• Image artifacts in OCT degrade quality, making manual assessment time-consuming and challenging.
• Automated quality control is essential for reliable OCT image analysis and clinical decision-making.

Purpose of the Study:

• To develop a novel, automated methodology for early quality assessment of OCT images.
• To differentiate between high-quality and poor-quality retinal OCT images efficiently.
• To enhance the reliability of data used in clinical diagnosis of retinal diseases.

Main Methods:

• Employed stochastic differential equations (SDE) to model the α-stable nature of OCT images.
• Utilized a fractional Laplacian filter to represent OCT image characteristics.
• Integrated α-stable parameters with a Support Vector Machine (SVM) for image quality classification.

Main Results:

• The proposed method demonstrated outstanding performance in detecting poor-quality OCT images.
• Validated on a large dataset of normal and abnormal retinal OCT images.
• The methodology shows applicability across various OCT scanning devices.

Conclusions:

• Automated quality control of retinal OCT images is feasible and highly effective using the proposed SDE-based method.
• This technique offers a reliable solution for identifying suboptimal images, reducing diagnostic errors.
• Clinical relevance lies in providing dependable data for diagnosing retinal and systemic diseases.