Skip to content

Creating Phantom Retinas for Fundus Imaging – ZEISS

Authors

Riley Semler, Shuo Wang, Isaac Ondo

Faculty Advisor

Mark Buckley

Abstract

When training clinicians to image human eyes, it is important to have accurate models for them to practice on and to calibrate the instruments by. Our team was tasked with creating long lasting, shelf stable retina phantoms, or model retinas, which accurately match the reflection spectrum and visual details of a human retina. These models for fundus imaging are to show several key features, such as blood vessels, the optic nerve, and the retinal disease drusen, along with accurately mimicking the pigments in the eye, such as melanin. We did this by doping contact lenses with various materials and dyes and stacking them to create the layers we wanted. The camera used for measurements was the handheld Zeiss Visuscout camera. In the end, we created several phantoms which replicated different structures and colors of the parts of the retina

Background

Having accurate models of the eye is important for both training doctors to use the company’s products as well as calibrating the devices. The company Zeiss asked our senior design group to look at possible ways to create these models and address several limitations that models currently on the market have, namely shelf life. These cameras use fundus imaging, which is a way of imaging the retina that only takes a picture inside the eye, and doesn’t see three dimensional depth of structures in the eye, like optical coherence tomography does.

The camera we were given to test with was the Zeiss Visuscout, which is a portable handheld camera, and has similar responses to wavelengths as their other products these phantoms would be used on, like the Clarus 500.

Figure 1. Graph of absorptivity of certain key pigments[1]

A close-up of a red round object

AI-generated content may be incorrect.

Figure 2. Picture of human retina with drusen obtained through fundus imaging [2]

A close-up of a device

AI-generated content may be incorrect.

Figure 3. Picture of Visuscout camera [3]

We wanted our phantoms to:

  • Simulate the images of the retinal structure of real eyes under a wide range of wavelengths.
  • Simulate the color of the retina
  • Simulate the blood vessels and optic nerve
  • Replicate disease patterns, such as Drusen

Results

We researched materials and created several kinds of phantoms. The method we chose of creating these was to dope contact lenses by leaving them to soak in a solution with the dying agent until they were saturated, and then to dry these out. We chose contact lenses as they had approximately the right curvature and thickness as the retina. We were not creating the whole housing system for the phantoms, so we took pictures of our phantoms inside the Zeiss OEMI model eye, with our phantoms placed in the back where the retina is.

A circular object with a black background

AI-generated content may be incorrect.
A close-up of a red round object

AI-generated content may be incorrect.

Figure 4. Comparison of our retina phantom’s representation of drusen (top) with a picture from published literature on drusen (bottom) [2]

A close-up of a circular object

AI-generated content may be incorrect.
A close-up of a human body

AI-generated content may be incorrect.

Figure 5. Comparison for the optic nerve of the phantom (top) and enlarged picture of the optic nerve taken with the Visuscout (bottom)

We ran into a problem midway through the semester where the camera could only get blurry photos in the OEMI eye model housing, so our latest pictures were not able to show the detail on our models properly.

A close-up of a red light

AI-generated content may be incorrect.

Figure 6. Images of the phantom retinas to show blood vessels

Conclusions

We were able to create several phantoms to mimic different systems in the eye. Additionally, choosing materials to represent different layers of the human retina takes significant trial and error in terms of material combinations, differing concentrations, and trying new materials all together. These imaging phantoms serve as a starting point for creating more fully fledged, shelf-stable products that companies use to help in testing for eye disease.

Acknowledgements / References

Funding:

This research was funded in part by the Zeiss Corporation and the University of Rochester Institute of Optics

References:

[1] Berendschot, J., DeLint, P., Norren, D. (2003). Fundus reflectance—historical and present ideas. Progress in Retinal and Eye Research 22 (2003) 171–200. doi: 10.1016/s1350-9462(02)00060-5. PMID: 12604057.

[2] Khan, J. (2010). The ABC’s of Drusen. Casey Eye Institute. [Accessed at https://www.ohsu.edu/casey-eye-institute/abcs-drusen].

[3] Carl Zeiss Vision. (n.d.). Fundus camera: Diagnose retinal and optic nerve conditions. ZEISS Vision Care. [Accessed at https://www.zeiss.com/vision-care/us/eye-care-professionals/equipment/eye-condition-check/fundus-camera.html].