How and Why to Scan All the Vertebrates ‐ Open Access Data as a Transformative Tool

Abstract

I will report on a large scale effort by over 50 investigators at 15 institutions to CT scan over 20,000 species of vertebrate. These data are made available ‘open access' on the MorphoSource website as they are gathered, with a CC-BY-NC copyright. This sprang from an ongoing initiative to scan all of the actinopterygian fishes (30,000+ species). The project, called “oVert - Open Exploration of Vertebrate Diversity in 3D” is directed by Dr. David Blackburn at the University of Florida. We are scanning museum specimens of nearly every genus of vertebrate with finer grained coverage in many areas. For example, Blackburn has already CT scanned a representative from every family of frog, and my lab has scanned at least one representative from 333 families of ray-finned fishes. We have developed protocols and tools to make mass scanning of specimens an effective strategy. The standard resolution for the project is expressed relative to the size of the animal - a minimum of 250 slices across the skull serves most purposes. Our slice data has been used to make models for 3D printing, new species descriptions, finite element modeling, digital animation, and 3D morphometrics. More than 2500 species of fishes and over 4000 specimens have been scanned at a single facility over the past 18 months. The peak scanning throughput was 740 specimens in a month. High throughput techniques and a scanner free of hourly charges has allowed investigators unprecedented ability to examine the ontogeny and population level diversity of skeletons. This has been accomplished without a scanning technician -- through a viral training program. Once the raw projection data are transformed into slice data, these can be analyzed using free, open source software programs such as Slicer, Drishti, ImageJ, and Horos. This greatly reduces barriers to dealing with these types of data by making it possible to perform the analysis remotely from the CT scanner. The immediate availability of the data to the community has already led to work not envisioned by the investigators gathering the data. A biologist in Germany wrote an interactive surface generator for the slice data we have posted, and he now hosts more than 3000 rotatable and printable 3D models of vertebrates. Segmentation of the data on a very large scale has been used to determine the 3D position and morphology of the six otoliths (ear bones) in the fish skull. Combined with habitat and life history data this will allow a look at the evolutionary pressures that led to otolith diversity. The general segmentation problem, isolating individual fishes from a mass of fish skeletons, inspired a new algorithm for processing landmark free data. In short, though these data have only been available for a short time we are already seeing the benefits of an open access policy. Support or Funding Information National Science Foundation (DBI-1701665)National Science Foundation (IOS- 1256602) A female Dellichthys morelandi, a New Zealand clingfish 35mm total length. Scanned at 14um resolution on a Bruker Skyscan 1173 with 9 other specimens. This abstract is from the Experimental Biology 2018 Meeting. There is no full text article associated with this abstract published in The FASEB Journal.