Abstract
Purpose
To remove the eye motion and stabilize the optical frequency domain imaging (OFDI) system for obtaining high quality images of the optic nerve head (ONH) and the pore structure of the lamina cribrosa.
Methods
An optical coherence tomography (OCT) instrument was combined with an active eye tracking system to compensate for eye motion in OCT imaging. The OCT system was a phase-stabilized deeply penetrating OFDI system operating at center wavelength of 1040 nm and the eye tracker was an 840 nm scanning laser ophthalmoscope (SLO). Retinal tracking was performed using real-time analysis of the distortions within SLO frames. OFDI had axial resolution of 4.8 µm (6.5 µm in air) and the theoretical spot-size on the retina was 13.7 µm. Eye motion was reported at a rate of 960 Hz and motion signals were inverted to correction signals and used to keep the OCT scanning grid locked on the same retinal area throughout the measurement. In the case of a tracking lock failure (e.g. blink or large saccade), the tracker signaled the OFDI system to rescan corrupted B-scans immediately stepping back 10 B-scans and holding the position until signal was valid again. The achieved tracking bandwidth was 32 Hz due to an internal time lag of the hardware. The ONH of a healthy volunteer was imaged over an area of 2.7 × 2.7 mm (8.8°) using 700 A-scans/B-scan. To visualize the benefit of the tracking, each acquired B-scan in a volume dataset (total of 700 B-scans) was integrated over depth to create an enface image of the ONH.
Results
The ONH was successfully imaged with negligible artifacts from eye motion (Fig. 1). On the left side, the whole dataset is seen including the duplicate corrupted B-scans. The corrupted B-scans were then removed in post-processing, thus leaving the undistorted duplicates untouched. The measured residual motion in the OCT B-scans was 0.32 arcmin (~1.6 µm) in a human eye. Four volumes from the same location were registered together to visualize the lamina cribrosa throughout the different depth slices of the eye (Fig. 2). The pore structure was clearly visible up to 430 um from the bottom of the ONH cup.
Conclusions
It is possible to obtain high quality OCT images from ONH and lamina cribrosa by compensating the eye motion during the measurements.
To remove the eye motion and stabilize the optical frequency domain imaging (OFDI) system for obtaining high quality images of the optic nerve head (ONH) and the pore structure of the lamina cribrosa.
Methods
An optical coherence tomography (OCT) instrument was combined with an active eye tracking system to compensate for eye motion in OCT imaging. The OCT system was a phase-stabilized deeply penetrating OFDI system operating at center wavelength of 1040 nm and the eye tracker was an 840 nm scanning laser ophthalmoscope (SLO). Retinal tracking was performed using real-time analysis of the distortions within SLO frames. OFDI had axial resolution of 4.8 µm (6.5 µm in air) and the theoretical spot-size on the retina was 13.7 µm. Eye motion was reported at a rate of 960 Hz and motion signals were inverted to correction signals and used to keep the OCT scanning grid locked on the same retinal area throughout the measurement. In the case of a tracking lock failure (e.g. blink or large saccade), the tracker signaled the OFDI system to rescan corrupted B-scans immediately stepping back 10 B-scans and holding the position until signal was valid again. The achieved tracking bandwidth was 32 Hz due to an internal time lag of the hardware. The ONH of a healthy volunteer was imaged over an area of 2.7 × 2.7 mm (8.8°) using 700 A-scans/B-scan. To visualize the benefit of the tracking, each acquired B-scan in a volume dataset (total of 700 B-scans) was integrated over depth to create an enface image of the ONH.
Results
The ONH was successfully imaged with negligible artifacts from eye motion (Fig. 1). On the left side, the whole dataset is seen including the duplicate corrupted B-scans. The corrupted B-scans were then removed in post-processing, thus leaving the undistorted duplicates untouched. The measured residual motion in the OCT B-scans was 0.32 arcmin (~1.6 µm) in a human eye. Four volumes from the same location were registered together to visualize the lamina cribrosa throughout the different depth slices of the eye (Fig. 2). The pore structure was clearly visible up to 430 um from the bottom of the ONH cup.
Conclusions
It is possible to obtain high quality OCT images from ONH and lamina cribrosa by compensating the eye motion during the measurements.
| Original language | English |
|---|---|
| Pages | 1452 |
| Number of pages | 1 |
| Publication status | Published - 5 May 2013 |
| Event | The Association for Research in Vision and Ophthalmology Annual Meeting 2013 - The Washington State Convention Center, Seattle, United States Duration: 5 May 2013 → 9 May 2013 http://www.arvo.org/Conferences_and_Courses/Past_Conferences/ |
Conference
| Conference | The Association for Research in Vision and Ophthalmology Annual Meeting 2013 |
|---|---|
| Abbreviated title | ARVO 2013 |
| Country/Territory | United States |
| City | Seattle |
| Period | 5/05/13 → 9/05/13 |
| Internet address |
Keywords
- optical coherence tomography
- retinal imaging
- scanning laser ophthalmoscope
- optic nerve head