Top Banner
Retinal Imaging Protocols for Constructing High Resolution Mosaics of In Vivo Photoreceptor Cells Blanca E. Marinez Cabrillo College Aptos, California Thomas Hebert, Austin Roorda, College of Optometry, University of Houston, Houston, Tx.
23

Blanc a E. Marine z Ca brillo College Ap tos, California

Jan 14, 2016

Download

Documents

Josh

Ret inal I maging Pr otocols for Constructin g High Resolution M osaics of In Vivo Photorec eptor Cells. Blanc a E. Marine z Ca brillo College Ap tos, California Th omas H ebert, Austin Ro orda , Col lege o f Optometry, Uni versity of Houston, Houston, Tx. - PowerPoint PPT Presentation
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Page 1: Blanc a E. Marine z Ca brillo College Ap tos, California

Retinal Imaging Protocols for Constructing High Resolution

Mosaics of In Vivo Photoreceptor Cells

Blanca E. MarinezCabrillo CollegeAptos, California

Thomas Hebert, Austin Roorda, College of Optometry, University of

Houston, Houston, Tx.

Page 2: Blanc a E. Marine z Ca brillo College Ap tos, California

Why is retinal imaging important?

Retinal imaging, the process of obtaining detailed images of the retina, it is important for:

• Early diagnosis of retinal diseases.

• Monitoring the progression of retinal diseases.

• Studying the distribution of photoreceptors throughout the retina.

AOSLO Image 1.5x1.4 deg

Courtesy: Austin Roorda

Page 3: Blanc a E. Marine z Ca brillo College Ap tos, California

How is retinal imaging possible?

Adaptive Optics Scanning Laser Ophthalmoscope (AOSLO)

wavefrontwavefrontsensingsensing

light light deliverydelivery

wavefrontwavefrontcompensationcompensation

lightlightdetectiondetection

rasterrasterscanningscanning eyeeye

Courtesy: Austin Roorda

Page 4: Blanc a E. Marine z Ca brillo College Ap tos, California

M O S A I C S

To Construct a mosaic…

Why is the protocol important?

High correlation between images is important when matching them.

THE B I G PICTURE…

Page 5: Blanc a E. Marine z Ca brillo College Ap tos, California

M O S A I C S

To Construct a mosaic…

Why is the protocol important?

High correlation between images is important when matching them.

THE B I G PICTURE…

Page 6: Blanc a E. Marine z Ca brillo College Ap tos, California

M O S A I C S

To Construct a mosaic…

Why is the protocol important?

High correlation between images is important when matching them.

THE B I G PICTURE…

Page 7: Blanc a E. Marine z Ca brillo College Ap tos, California

M O S A I C S

To Construct a mosaic…

Why is the protocol important?

High correlation between images is important when matching them.

THE B I G PICTURE…

Page 8: Blanc a E. Marine z Ca brillo College Ap tos, California

M O S A I C S

To Construct a mosaic…

Why is the protocol important?

High correlation between images is important when matching them.

(Images from different protocols)

The protocol affects the clarity of the images, the ability of the subject to fixate well, the region of the retina being imaged and the problems when creating the mosaic.

THE B I G PICTURE…

Page 9: Blanc a E. Marine z Ca brillo College Ap tos, California

AOSLO Image 1.5x1.4 deg

Courtesy: Austin Roorda

This shows a mosaic of approximately 9x9 degrees co-aligned with the same region of the fundus photograph.

Page 10: Blanc a E. Marine z Ca brillo College Ap tos, California

Targets allow you to map the region of the retina that you want to image and to keep track of where your images should match when constructing the mosaic.

P r o t o c o lIt’s what you look at…

how you look at it...

And how we collect the data.

1

5

16 6

7

15

8

9

10

11

12

13

142

3

4

1

5

16 6

7

15

8

9

10

11

12

13

142

3

4

0

40

100110

120

130

140

150

160

170

180

190

200

210

220

230240

250260 270 280

290300

330

50

310

6070

340

320

90 80

350

10

20

30

Page 11: Blanc a E. Marine z Ca brillo College Ap tos, California

P r o t o c o lIt’s what you look at…

how you look at it...

And how we collect the data.

The subject can fixate at one point of the target, gaze across given points or line on the target, or follow a moving point of light like a laser.

Video of a subject gazing across a line on a target.

Page 12: Blanc a E. Marine z Ca brillo College Ap tos, California

P r o t o c o lIt’s what you look at…

how you look at it...

And how we collect the data.

The subject can fixate at one point of the target, gaze across given points or line on the target, or follow a moving point of light like a laser.

Video of a subject fixating at one point of the target.

Page 13: Blanc a E. Marine z Ca brillo College Ap tos, California

P r o t o c o l It’s what you look at…

how you look at it...

And how we collect the data.

• The length of your videos.You may vary:

•The number of images you register and sum for each location.

• The number of videos you take.

Page 14: Blanc a E. Marine z Ca brillo College Ap tos, California

Protocol A

Target

1

5

16 6

7

15

8

9

10

11

12

13

142

3

4

Twenty-five videos, twenty seconds each.

Page 15: Blanc a E. Marine z Ca brillo College Ap tos, California

Protocol A

•Drastic difference in intensity of photoreceptors in the two images.

•Low correlation between images.

•Edges are more prominent.

Courtesy: Ramesh Sundaram

KV

Page 16: Blanc a E. Marine z Ca brillo College Ap tos, California

Protocol B

1

5

16 6

7

15

8

9

10

11

12

13

142

3

4

Target

Ten continuous videos where the subject gazes across given points.

Page 17: Blanc a E. Marine z Ca brillo College Ap tos, California

Protocol B

RS

• Noticeable change in images from different videos.

•Good foveal region due to the amount of videos taken at the fovea.

•When constructing the mosaic there was a high correlation between images of the same video but not between corresponding images of different videos.

Page 18: Blanc a E. Marine z Ca brillo College Ap tos, California

Protocol C

Target

One continuous video where the subject gazes across lines 1-11.

Page 19: Blanc a E. Marine z Ca brillo College Ap tos, California

Protocol C

RS

•The mosaic shows uniform brightness.

•Edges are not prominent.

•High correlation between images.

•The subject had difficulties gazing across the lines.

•Fixation on the central region was minimal therefore we didn’t get a good image of the foveal region.

Page 20: Blanc a E. Marine z Ca brillo College Ap tos, California

Protocol D

Target

One continuous video where the subject followed a laser point move along lines 1-11.

Page 21: Blanc a E. Marine z Ca brillo College Ap tos, California

Protocol D

RS•The mosaic shows uniform brightness.

•Edges are not prominent

•High correlation between images.

•Subject could fixate well following the laser across the lines.

• Fixation on the central region was minimal therefore we didn’t get a good image of the foveal region.

Page 22: Blanc a E. Marine z Ca brillo College Ap tos, California

C o n c l u s i o n

Subjectively, the best results were given by Protocol D. This protocol allows the subject to fixate well and it minimizes the effect of change in reflectivity of photoreceptors due to time difference in which the images were captured. It also allows us to match common features of neighboring frames more accurately. This creates a mosaic that is uniform in brightness, that has less visible edges, and easy to construct.

Page 23: Blanc a E. Marine z Ca brillo College Ap tos, California

Acknowledgments

• Center for Adaptive Optics• National Science Foundation #AST-987683• Dr. Tom Hebert• Dr. Austin Roorda• Dr. Fernando Romero• Ramesh Sundaram• Dr. Krishnakumar Venkateswaran • Siddharth Poonja• The Roorda Lab