 |
|
|
Description of the Adaptive Optics Flood Illumination Camera
Figure 1 shows a schematic and photograph of the Indiana AO flood-illumination retina camera. The camera consists of three sub-systems:
A dental impression attached to a sturdy xyz bite bar translation stage stabilizes the head and provides accurate pupil positioning.
High-Speed Conventional Flood-illumination Imaging: A 200 mW multimode laser diode (670 nm) coupled to 300 m of multimode fiber uniformly illuminates a 0.8 degree to 1.8 degree patch of retina. In addition a 17 mW SLD (835 nm) coupled to 25 m of multimode fiber uniformly illuminates a 0.5 degree patch of retina. A back-illuminated scientific CCD (Quantix, Roper Scientific, Inc.) captures aerial images of the retina through a 6 mm pupil, and its image acquisition is synchronized to the strobing fiber light sources. Exposure duration is set between 1/10th ms and 10 ms. Maximum frame rates are 30 Hz (0.8 degree field) and 60 Hz (0.4 degree field) for continuous imaging and 500 Hz (0.5 degree x1 degree field) for short-burst imaging.
AO Wavefront Correction: The main components of the AO system are a 37 actuator Xinetics deformable mirror and a Shack-Hartmann wavefront sensor employing a 17 x 17 lenslet array. A 788 nm SLD is used as the beacon. The closed-loop control performs up to 22 wavefront measurements and corrections per second. Lower rates are often used as they permit additional AO diagnostic tools to execute in the background. Measurement and correction is across a 6.8 mm pupil.
   |
| Figure 1. (top) Detailed layout of the AO flood-illumination retina camera. The camera consists of three channels: (1) sample channel, (2) reference channel, and (3) detection channel. (middle) (red line) The AO system corrects the ocular aberrations using a 788 nm SLD, Shack-Hartmann wavefront sensor, and 37-actuator Xinetics deformable mirror. (green line) Conventional flood-illumination is realized with several different laser and superluminescent diodes that emit at visible and near-infrared wavelengths, and can operate simultaneously. (bottom) A more recent photograph of the camera and group members Jungtae, Weihua, Barry, Yan, and Ravi (from left to right). |
AO Flood-Illumination Results
Figure 2 shows single, one millisecond snapshots of the cone mosaic for one subject with and without AO correction. The images illustrate the substantial gain in contrast and clarity afforded by correcting the most significant wave aberrations of the eye beyond those of sphere and cylinder refractive error.
 |
 |
| Figure 2. Raw flood illuminated images of the cone mosaic centered at 1.25 degrees eccentricity in one subject's eye without (left) and with (right) adaptive compensation. For both images, best correction of defocus and astigmatism was achieved with trial lenses and axial translation of the science CCD camera. The 1-degree patch of retina was illuminated at 679 nm. Small bright spots depict light reflecting from individual photoreceptor cells. The fuzzy dark line is a shadow of a 7 µm diameter capillary lying in an out-of-focus plane. | |
The two AO compensated, 30 Hz, videos in Figure 3 below show the cone mosaic and a through focus of the retina. In the left video, individual cone cells fill the 0.8-degree field of view. Video rate imaging reveals the extent of the retina motion as the subject fixates at the same point. The right video depicts a through focus of the retina that was acquired starting at the retinal vasculature and ending at the photoreceptor layer. The actual figure shows four frames corresponding to the anterior retina with retinal capillaries in focus, intermediate depths, and with cone photoreceptors in focus. For this video, custom registration software effectively removed retina motion artifacts, which were comparable in magnitude to that in Figure 3 (left). Registration permits improved visual and quantitative analysis of videos.
The AO compensated 60 Hz videos in Figure 4 show small 0.4-degree patches of the cone mosaic and retinal vasculature. Signal to noise is essentially the same as for the 30 Hz (Fig. 3) video as retinal illumination, exposure duration, and CCD read noise did not change. In the left video, individual cone cells fill the 0.4-degree field of view. The right video reveals cellular flow dynamics inside a vessel and adjacent 6 to 7 µm capillary.
Figure 5 shows representative four-burst videos acquired at 500 Hz of the same retinal capillaries in one subject with and without adaptive compensation. In the uncompensated video (top), the capillaries are highly blurred due to the relatively poor image quality, though gross blood motion (flow is from left to right in the video) can be detected. Correction of the most significant aberrations in the eye (bottom) reveals a complex capillary network (6 to 7 µm diameter) with a large vessel located in the upper right corner. The vessels are characterized by a flowing stream of light and dark reflections that are suggestive of chains of erythrocytes or plasma gaps. Blood velocity in the capillaries was directly measured at 1.5 mm/s. In general, short-burst imaging provides sufficient temporal bandwidth, sensitivity and optical resolution to clearly capture high temporal dynamics of the microscopic retina, in this case the motion of single blood cells in capillaries.
 |
|
 |
 |
 |
 |
|
|
|
|
 |
| |
|
|
|
anterior | intermediate depths | posterior | |
|
Figure 3. Single frame of the cone mosaic (left) and sequence of frames of a through focus of the retina (right) extracted from raw 30 Hz flood illuminated videos with dynamic adaptive compensation. The illumination patch subtended 0.8 degree x 0.8 degree (left) and 0.67 degree x 0.57 degree (right) at 1.4 degrees (left) and 1.25 degrees (right) eccentricity. Adaptive compensation occurred at 15 Hz (left) and 22 Hz (right). Exposure duration is 2 (left) and 1 (right) ms. Illumination was at 670 nm. Scale bars represent 50 µm.
Download Figure 3 (left) video: |
|||||
 |
|
 |
|
Figure 4. (left) First frame of a 60 Hz flood illuminated video of the cone mosaic with dynamic adaptive compensation. The illumination patch subtended 0.4 degree at 1.4 degrees retinal eccentricity. Exposure duration is 2 ms. (right) Sequence of raw single images from a 60 Hz flood illuminated video that depicts the flow of individual blood cells in a retinal capillary. The illumination patch subtended 0.4 degree at 2.5 degrees. Exposure duration is 1 ms. Dark arrows point to a single blood cell whose brightness fluctuates as it traverses the capillary. The time duration between the first and fourth images is 50 ms. Illumination was at 670 nm. Scale bars represent 25 µm (left) and 60 µm (right). Click to download videos (1.0 MB, 0.5 MB).
Download Figure 4 (left) video: |
||
 |
Figure 5. First frame of four-burst videos without (top) and with (bottom) adaptive compensation of a network of retinal capillaries at 1.4 degrees eccentricity in one subject. The capillaries are the smallest in the retina, being 6 to 7 µm in diameter. The size of the retinal patch is 1 degree by 1/2 degree. Both videos were captured at 500 Hz using a 1 ms exposure and 1 ms delay.
Download Figure 5 video: |
In the previous cone videos above (in particular, Figure 4 left), the reflectance of some isolated cones fluctuate, in some cases disappearing and then reappearing within a second or two. We have observed that the rapidity and acuteness of the fluctuations and number of cones that fluctuate in a given patch of retina depend on the temporal coherence of the light source and the bleached state of the retina, with fully regenerated cones (dark adapted eye) providing the most variation. Figure 6 below is an example of a dark adapted eye. Intensity of the 670 nm flashes during the video is sufficient to fully bleach both the medium (M) and long (L) wavelength cones in roughly the first second. Fluctuations are clearly much stronger and widespread than in the previous videos. Interestingly, the cones fluctuate at distinctly different temporal periods, with some highly stable throughout the video (no fluctuations) and others oscillating quickly at the beginning, and then gradually slowing and stabilizing as the video progresses. To compare temporal differences between cones, the right side of Figure 6 shows three extracted cones that are representative of the three general groups of oscillations in the video. Cone 1 oscillates quickly, going through 1.5 oscillations in the first second (30 frames) with no appreciable oscillations afterwards. Cone 2 oscillates more slowly, and cone 3 does not oscillate at all. These three groups of oscillations might correspond to the three spectral subtypes of cones, which bleach at different rates when illuminated with the flashing 670 nm light. In general, the intensity fluctuations observed here, which are indicative of structural changes occurring inside individual cone cells, are readily revealed by the high speed acquisition and high sensitivity of the Indiana AO flood illumination camera.
 |
 |
|
Figure 6. First frame (left) and sequence of frames of three individual cones (right) extracted from a video acquired at 30 Hz in a dark adapted eye. Imaging/bleaching is at 670 nm (Δλ = 0.8 nm); exposure duration is 2 ms. The imaging light, which flashes at 30 Hz, also acts as the bleaching source, bleaching the photopigment during each exposure. The total time duration for the 21-frame sequence is 700 ms.
Download Figure 6 video: |
|
|
URL: http://www.opt.indiana.edu/people/faculty/miller/aoflood.htm Revised: March 20, 2007 | |
 |
IU Optometry home page: http://www.opt.indiana.edu/ Comments: Web Administrator |
 |
Copyright © 2007 The Trustees of Indiana University | Copyright Complaints |