A Computer Model of the Retina Reveals Some of the Eye's Architecture

An international team of researchers has used advanced computer modeling to discover how one type of cell in the retina can shape the retinal output in a variety of ways, paving the way for a better understanding of eye diseases and their potential outcomes.

The group combined work in genetics and computer modeling to build a model that predicted the changes observed to the visual output when horizontal cells are deactivated, according to the resultant study.

In the retina, photoreceptors are the input channel that take in the light, ganglion cells are the output channel that send the information to the brain, and in between, other types of cells, including horizontal and bipolar cells, mediate the process.

>> READ: The Silicon Valley of Health Analytics Is Taking on Diabetic Retinopathy

In the first stage, a team from the Friedrich Miescher Institute (FMI) in Basel deactivated the horizontal cells in mouse retinas chemogenetically and observed the effect on how ganglion cells responded and processed the same visual input before and after.

“Surprisingly, the perturbation caused a large set of different changes in the output of the retina,” said Rava da Silveira, PhD, from the Ecole Normale Supérieure in Paris, one of the paper’s senior authors. Rather than one effect, the perturbation the horizontal cells produced six, suggesting a more complicated architecture than a simple on-off relationship. “By turning one knob in the retina, you change different output channels in different ways,” da Silveira said.

The effects included sharpening responses, allowing for delayed response to light, and acted in distinct ways on the roughly 30 types of ganglion cells.

After the effects were observed, researchers from ETH Zurich and the Ecole Normale Supérieure built a computer model of the retina to simulate the same process. The model successfully predicted all the effects observed in the FMI experiment and made additional predictions about the role of horizontal cells, findings that were not initially observed in the data.

The additional predictions were tested and found to be accurate, according to Felix Franke, PhD, a bioengineering professor at ETH Zurich and co-author of the paper.

While the work will have few immediate effects on the treatment of eye diseases, it’s one step toward a better understanding them. When treating diseases, experts often know what’s wrong but don’t understand the cascade of events that are causing the problem. The computer model, however, can open a window into the retinal circuitry.

“The model not only helps you capture what’s going on,” said da Silveira. “It’s also telling you how this happens.”

Get the best insights in healthcare analytics directly to your inbox.

Related

The Slow, Frustrating Rise of the Electronic Health Record

How to Minimize Data Chaos in Healthcare

Employing Eye Experts, Google Sharpens Its Disease-Detecting AI