Correcting for physical distortions in visual stimuli improves reproducibility in zebrafish neuroscience

Timothy Dunn, James E. Fitzgerald

Abstract

Breakthrough technologies for monitoring and manipulating single-neuron activity provide unprecedented opportunities for whole-brain neuroscience in larval zebrafish. Understanding the neural mechanisms of visually guided behavior also requires precise stimulus control, but little prior research has accounted for physical distortions that result from refraction and reflection at an air-water interface that usually separates the projected stimulus from the fish. Here we provide a computational tool that transforms between projected and received stimuli in order to detect and control these distortions. The tool considers the most commonly encountered interface geometry, and we show that this and other common configurations produce stereotyped distortions. By correcting these distortions, we reduced discrepancies in the literature concerning stimuli that evoke escape behavior, and we expect this tool will help reconcile other confusing aspects of the literature. This tool also aids experimental design, and we illustrate the dangers that uncorrected stimuli pose to receptive field mapping experiments.

Introduction

In zebrafish neuroscience, researchers often project visual stimuli onto a screen outside a water-filled dish. However, the air-water interface between the screen and the fish introduces significant optical distortions due to refraction (Snell’s Law) and reflection (Fresnel equations). Despite the high precision of neural imaging, many studies treat "projected" stimuli as identical to "received" stimuli. This "naive" approach can lead to major reproducibility issues, as different experimental geometries (e.g., flat vs. curved dishes) produce distinct optical artifacts, potentially leading to conflicting biological conclusions.

Methods

We developed a mathematical framework to map the relationship between the projected position of a stimulus ($\theta'$) and its apparent position as perceived by the fish ($\theta$).

Experimental setup and optical modeling showing the relationship between projected and perceived stimuli.

The framework models the two most common experimental configurations:

  1. Flat interfaces: Stimuli projected through a flat tank bottom.
  2. Curved interfaces: Stimuli projected through a cylindrical dish.

By treating the fish eye as a pinhole detector and tracing light paths through air, plastic, and water, we derived trigonometric solutions for spatial mapping. Additionally, we used Fresnel equations to model how light transmittance—and thus perceived contrast—drops at steeper angles. We have released an open-source tool to allow researchers to "pre-distort" stimuli or "post-correct" existing data.

Results and Insights

Reconciling Behavioral Discrepancies

The model was validated using two landmark datasets on looming-evoked escape behavior: Temizer et al. (2015), which used a curved interface, and Dunn et al. (2016), which used a flat interface. These studies reported wildly different "critical angles" for triggering escapes ($21.7^\circ$ vs. $72.0^\circ$). By applying our corrections and calculating the true solid angle on the eye, the $50.3^\circ$ angular discrepancy between these findings was reduced to a statistically insignificant difference.

The Snell Window and Transmittance

Refraction compresses the entire $180^\circ$ external horizon into a cone of approximately $97.2^\circ$ from the fish’s perspective, known as the Snell window. Stimuli near the edges of this window are severely "squashed." Furthermore, transmittance remains high centrally but drops sharply toward the edges, creating a "vignetting" effect that impacts contrast sensitivity measurements.

Receptive Field Mapping

Simulations demonstrate that failing to account for Snell’s Law can lead to substantial errors in receptive field mapping. For neurons sensitive to the periphery, uncorrected data can result in significant misestimates of both the position and size of neural receptive fields.

Discussion

While our model provides a vital correction tool, it currently employs a pinhole eye model and assumes a fixed fish position relative to the interface. In free-swimming assays, the changing distance ($d_w$) presents an ongoing engineering challenge for real-time correction. Nevertheless, this work highlights that ignoring basic environmental physics can compromise the accuracy and reproducibility of zebrafish neuroscience.

Summary updated

Abstract

Breakthrough technologies for monitoring and manipulating single-neuron activity provide unprecedented opportunities for whole-brain neuroscience in larval zebrafish 1–9 . Understanding the neural mechanisms of visually guided behavior also requires precise stimulus control, but little prior research has accounted for physical distortions that result from refraction and reflection at an air-water interface that usually separates the projected stimulus from the fish 10–12 . Here we provide a computational tool that transforms between projected and received stimuli in order to detect and control these distortions. The tool considers the most commonly encountered interface geometry, and we show that this and other common configurations produce stereotyped distortions. By correcting these distortions, we reduced discrepancies in the literature concerning stimuli that evoke escape behavior 13,14 , and we expect this tool will help reconcile other confusing aspects of the literature. This tool also aids experimental design, and we illustrate the dangers that uncorrected stimuli pose to receptive field mapping experiments.
BibTeX
@inproceedings{dunn2020correcting,
  author = {Timothy Dunn and James E. Fitzgerald},
  title = {{Correcting for physical distortions in visual stimuli improves reproducibility in zebrafish neuroscience}},
  booktitle = {eLife},
  year = {2020},
  doi = {10.7554/elife.53684},
}
Figure