Related Experiment Video
Updated: Mar 26, 2026
08:17
Optical Coherence Tomography: Imaging Mouse Retinal Ganglion Cells In Vivo
Published on: September 22, 2017
20.4K
Segmentation Errors in Macular Ganglion Cell Analysis as Determined by Optical Coherence Tomography
Young Hoon Hwang1, Min Kyung Kim1, Dai Woo Kim2
1Department of Ophthalmology, Konyang University, Kim's Eye Hospital, Myung-Gok Eye Research Institute, Seoul, Korea.
Ophthalmology
|February 9, 2016
Summary
Optical coherence tomography (OCT) can have segmentation errors in macular ganglion cell inner plexiform layer (GCIPL) thickness measurements, particularly in myopic eyes. Approximately half of these errors were not reproducible in repeat scans.
Area of Science:
- Ophthalmology
- Medical Imaging
- Retinal Imaging
Background:
- Accurate measurement of macular ganglion cell inner plexiform layer (GCIPL) thickness is crucial for diagnosing and monitoring retinal diseases.
- Optical coherence tomography (OCT) is a primary imaging modality for assessing GCIPL thickness.
- Segmentation errors in OCT can impact the reliability of GCIPL thickness measurements.
Purpose of the Study:
- To determine the prevalence, characteristics, contributing factors, and reproducibility of segmentation errors in OCT-derived macular GCIPL thickness measurements.
- To identify factors associated with the occurrence of these segmentation errors.
- To assess the consistency of segmentation errors across repeated OCT scans.
Main Methods:
- A cross-sectional study involving 538 eyes (290 subjects) with OCT-measured macular GCIPL thickness.
- Exclusion of eyes with macular disorders.
- Detailed inspection of OCT B-scan images to identify segmentation errors (presence, layer, location, area).
- Analysis of factors (age, refractive error, visual field, RNFL thickness, OCT signal strength) associated with errors.
- Repeat OCT examinations to evaluate error reproducibility.
Main Results:
- Segmentation errors were observed in 9.7% (52/538) of eyes.
- Common error features included involvement of both borders, nasal location, and diffuse nature.
- Higher myopia (P < 0.001) was significantly associated with segmentation errors in multivariate analysis.
- Segmentation errors were reproducible in only 46.2% of repeated examinations.
Conclusions:
- OCT segmentation algorithms can produce errors in macular GCIPL thickness measurements, particularly in eyes with high myopia.
- A significant proportion of these segmentation errors are not reproducible, highlighting potential variability in measurements.
- Clinicians should be aware of these potential segmentation errors and their influencing factors when interpreting OCT-based GCIPL thickness data.
