Fractal-based analysis of optical coherence tomography data to quantify retinal tissue damage

Area of Science:

• Ophthalmology
• Medical Imaging
• Biomedical Engineering

Background:

• Optical Coherence Tomography (OCT) assesses retinal morphology for early neural loss.
• Data from 74 healthy eyes and 43 with mild diabetic retinopathy (MDR) were analyzed.
• Custom algorithm (OCTRIMA) measured intraretinal layer thickness and fractal dimension.

Purpose of the Study:

• To evaluate OCT sensitivity in detecting early neural loss in diabetic retinopathy.
• To compare fractal dimension analysis with standard thickness measurements for diagnosing MDR.
• To identify retinal layers most susceptible to initial damage in MDR.

Main Methods:

• OCT images analyzed using OCTRIMA for intraretinal layer thickness.
• Power spectrum method calculated fractal dimension in retinal regions of interest.
• ANOVA, Newman-Keuls post-hoc tests, and ROC curves used for statistical analysis.

Main Results:

• Fractal dimension was higher in MDR eyes across most retinal layers compared to healthy eyes.
• Fractal dimension of the GCL+IPL complex showed the highest diagnostic accuracy (AUROC=0.96).
• Fractal analysis of the GCL+IPL complex demonstrated superior diagnostic capability for early DR compared to thickness measurements.

Conclusions:

• The GCL+IPL complex, OPL, and OS are most vulnerable to initial damage in MDR.
• Fractal analysis offers higher sensitivity for detecting early retinal neurodegeneration.
• Fractal analysis shows potential as a diagnostic predictor for early diabetic retinopathy.