Measuring Compound Eye Optics with Microscope and MicroCT Images

Abstract

The arthropod compound eye is the most prevalent eye type in the animal kingdom, with an impressive range of shapes and sizes. Studying its natural range of morphologies provides insight into visual ecology, development, and evolution. In contrast to the camera-type eyes we possess, external structures of compound eyes often reveal resolution, sensitivity, and field of view if the eye is spherical. Non-spherical eyes, however, require measuring internal structures using imaging technology like MicroCT (μCT). μCT is a burgeoning 3D X-Ray imaging technique that has been used to image arthropod muscles, brains, ocelli, and eyes. Thus far, there is no efficient tool to automate characterizing compound eye optics in either 2D or 3D data. We present two open-source programs: (1) the ommatidia detecting algorithm (ODA), which automatically measures ommatidia count and diameter in 2D images, and (2) a μCT pipeline (ODA-3D), which calculates anatomical acuity, sensitivity, and field of view across the eye by applying the ODA to 3D data. We validate these algorithms on images, images of replicas, and μCT scans from eyes of ants, fruit flies, moths, and a bee.

Acuity and summation strategies differ in vinegar and desert fruit flies

“…We further analyzed ommatidia distribution across the visual field to approximate parameters of optical performance. Ommatidial counts and diameters were measured using the ommatidia detecting algorithm (ODA; Currea et al., 2021 ), which measures the number and diameter, D , of corneal lenses by leveraging the periodic nature of the ommatidial lattice through the use of a low-pass filter and the 2D Fourier transform. Corneal lens area, which as opposed to diameter is directly proportional to optical sensitivity, was measured as the area of a circle defined by the facet diameter, A = π ( D /2) 2 .…”

“…To measure optically relevant parameters, an open source Python program ( Currea et al., 2021 ) generated a single image composite of the stack of photos ( Figure 3 A, bottom). Ommatidial counts and diameters were measured using the ommatidia detecting algorithm ( Currea et al., 2021 ), which measures the number and diameter, D , of ommatidia by using the fast fourier transform to generate the reciprocal image of the focus stack image ( Figure S2 ).…”

“…To measure optically relevant parameters, an open source Python program ( Currea et al., 2021 ) generated a single image composite of the stack of photos ( Figure 3 A, bottom). Ommatidial counts and diameters were measured using the ommatidia detecting algorithm ( Currea et al., 2021 ), which measures the number and diameter, D , of ommatidia by using the fast fourier transform to generate the reciprocal image of the focus stack image ( Figure S2 ). In the reciprocal image, each pixel represents a sinusoidal grating in the original image with a particular contrast (the pixel’s value), spatial frequency (its reciprocal distance from the center) and orientation (its polar angle).…”

“…By finding the fundamental frequencies of the reciprocal image (maxima closest to the center), the ODA applies a low-pass filter to the original image to more easily detect the center of ommatidia ( Figure S3 ). The ODA has been validated on the eyes of D. melanogaster , generating facet diameters and counts equal to 105 ± 3% and 104 ± 4% and with correlations of .93 and .82 of those measured manually ( Currea et al., 2021 ). The spherical model was also a good fit and explained roughly the same proportion of the variance in eye surface values (R 2 ≅ .8) for all three genotypes (plotted in Figure 3 D).…”

“… All data and code for generating the figures and statistics used in this study are publicly available on figshare.com, with DOIs listed in the key resources table . The Compound Eye Tools API used for detecting ommatidia in an image is publicly available on github.com ( https://github.com/jpcurrea/eye_tools/ ) and detailed in a preprint ( Currea et al., 2021 ). Any additional information required to reanalyze the data reported in this paper is available from the lead contact upon request.…”