Related Experiment Video
Updated: Jun 9, 2026
09:17
Using Retinal Imaging to Study Dementia
Published on: November 6, 2017
Improving image segmentation performance and quantitative analysis via a computer-aided grading methodology for
Delia Cabrera Debuc1, Harry M Salinas, Sudarshan Ranganathan
1University of Miami, Miller School of Medicine, Bascom Palmer Eye Institute, Miami, Florida 33136, USA. dcabrera2@med.miami.edu
Journal of Biomedical Optics
|August 31, 2010
Summary
A new computer-aided grading method improves quantitative analysis of optical coherence tomography (OCT) retinal images. This validated methodology enhances the accuracy and reproducibility of retinal thickness measurements in healthy eyes.
Area of Science:
- Ophthalmology
- Medical Imaging
- Biomedical Engineering
Background:
- Optical coherence tomography (OCT) is crucial for retinal imaging.
- Accurate quantitative analysis of OCT images is essential for diagnosing and monitoring eye diseases.
- Existing automated methods may have limitations in precision and error correction.
Purpose of the Study:
- To develop and validate a computer-aided grading methodology for quantitative analysis and error correction of OCT retinal images.
- To assess the reproducibility and accuracy of the developed methodology compared to standard automated OCT analysis.
- To evaluate the performance of the OCT retinal image analysis (OCTRIMA) algorithm.
Main Methods:
- A custom-built, computer-aided grading methodology was developed.
- Sixty Stratus OCT B-scans from ten healthy eyes were analyzed by two independent graders.
- Intergrader and intragrader reproducibility were assessed using Bland-Altman plots and Pearson correlation coefficients.
- Retinal thickness was compared between the custom methodology and automated Stratus OCT results.
Main Results:
- The custom methodology showed high intergrader and intragrader reproducibility (median thickness differences <1% and <2% of mean thickness).
- Measurement accuracy ranged from 0.27 to 1.76 microm for reproducibility tests.
- High correlation (R(2)>0.98) was observed between the custom methodology and Stratus OCT results across all ETDRS regions.
- Mean thickness differences varied depending on the chosen outer retinal border (IS/OS vs. OS/RPE junction).
Conclusions:
- The developed computer-aided grading methodology offers robust and localized quantification of retinal structure.
- This approach enhances the accuracy and reliability of OCT image analysis for both healthy controls and patients.
- The OCTRIMA algorithm demonstrates significant potential for improving clinical diagnosis and research in ophthalmology.
