Sue Goes Virtual
Art Andersen, Ralph E. Chapman and Linda T. Deck
Issue: Volume 33 Issue 7: (July 2010)

Sue Goes Virtual

There are many challenges to studying the biology of dinosaurs. Their skeletons are extremely fragile and, most often, incomplete. Because the bones are fossils, they may be distorted and usually composed of materials that were very different from what they were made of when living. And, dinosaurs are mainly quite large and physically difficult to study.

Yet, so many of the questions researchers need answers to are tied up in the bones and how they fit together to make skeletons. They include relatively simple matters, such as how much dinosaurs weighed, what they looked like fleshed out, how they moved, and even what their basic posture was. These questions must be answered before paleontologists can progress to addressing more complex problems. So, how do you minimize or circumvent the challenges and start answering these basic questions? Technology holds many of the answers.

These were problems facing Peter Makovicky, dinosaur curator and chair of the Department of Geology at the Field Museum of Natural History in Chicago. He wanted to start detailed biological studies of Sue, the Tyrannosaurus rex, as part of the museum’s celebration of the 10th anniversary of Sue’s unveiling. Makovicky turned to us­—Ralph Chapman and Linda Deck (Deck & Chapman LLC and New Mexico Virtualization of Los Alamos, New Mexico) and Art Andersen (Virtual Surfaces of Mt. Prospect, Illinois)—because we created the first real digital dinosaur during the 1990s based on the Smithsonian Institution’s Triceratops skeleton (see “No Bones About It,” February 2001).

Chapman and Deck have continued exploring the usefulness of 3D technology for museums since that project, and Andersen went on to apply 3D scanning technology to the study of the mummified duck-billed dinosaur known as Leonardo (see “Dino Might,” July 2009) as well as many other 3D projects.

Peter Makovicky (right) of the Chicago Field Museum, along with a team of scanning specialists, hold a research-grade cast of a femur from Sue, the museum’s T. rex, before CT-scanning it in order to make a high-res 3D model.

Sue presented all the classic challenges associated with dinosaur skeletons, magnified by the fact that the skeleton is huge even for a large meat-eating dinosaur. The original bones are incorporated into the mounted skeleton—in the main exhibit hall of the Field Museum—and cannot be removed easily for further work. On the plus side, the skeleton is unusually complete, and relatively few of those bones are not original. This was the situation we encountered upon entering the project; our task was to address all these conditions and generate a highly-detailed, complete skeleton of the T. rex.

Learning from the Past

Luckily, we had encountered this situation before and knew how to approach the work, accomplish the task, and keep Sue safe and continuously on display; visitors to the Field Museum would be disappointed if she was unavailable for viewing during their trip. We learned many lessons from virtualizing the Smithsonian’s Triceratops and used the solutions we came up with for that project—and, in some instances, improvements on them—to facilitate Sue’s entry into the virtual world. The project was done in three phases.

The first phase was to capture the whole skeleton in 3D at a relatively low resolution. We did this because the configuration of the mount of Sue was an awesome accomplishment by the Field Museum staff, which evaluated the bones and determined how they should go together in order to put the skeleton in a reasonable posture. Accomplishing this just as well virtually is still far off in the future because it is a very labor-intensive and hands-on process. So the skeleton as mounted represented a great resource that we wanted to use. We decided we must capture the whole thing—11 feet tall and more than 40 feet long.

To do this, we enlisted the Chicago Police Department (CPD), which had recently purchased a large-format Leica area laser scanner. They will be using the device for crime and accident investigations to reconstruct scenes in 3D. Andersen contacted Herb Keeler, the person in charge of the police scanning, and asked if they would be interested in working on Chicago’s own T. rex. The department decided that scanning the mount would be a good training exercise and that it could generate great publicity for the CPD.

Several casts of Sue’s vertebrae are CT-scanned at Loyola Medical Center. Larger bone casts had to be scanned at Ford.

Last fall, the CPD arrived at the Field Museum with the scanner, accessory items, and yellow crime-scene tape. The museum provided several large monitors so the public and press could view the data acquisition as it happened. Museum staff, the police representatives, and the three of us walked around and answered questions raised by that day’s visitors, and the atmosphere was festive. A total of seven scans were made, covering more than 90 percent of the mount, which allowed us to construct a full virtual copy of it. This model already is being used by researchers to study the animal. 

Studying Sue

Access to accurate digital models of fossils has revolutionized the study of biomechanics and functional morphology of extinct animals. Techniques such as laser and white-light scanning can capture the complex external anatomy of fossils in three dimensions, while CT scanning supplements those by affording internal views of inaccessible parts, such as braincases and marrow cavities. Combined with the powers of CAD software, paleobiologists can now attempt realistic digital models of fossil organisms to estimate size and mass, or determine parameters such as running speeds or bite forces.

A major scanning project to generate an accurate 3D model of Sue’s skeleton was initiated in conjunction with the celebration of the T. rex’s 10th anniversary of being displayed at the Field Museum. Our goals were threefold: attempt to model Sue’s mass from 3D models of her mounted skeleton and scans of individual bones, use those models to analyze agility, and develop accurate scale versions of bones for research and commercial purposes. In this regard, Sue is a superlative candidate. Her skeleton is not only the largest and most complete

T. rex, but the largest skeleton of a biped anywhere. As such, she defines a biological extreme. Given Sue’s size, the scanning project presented some technical challenges that were overcome by combining a variety of techniques.

Based on accurate scans of mounted skeletons, body outlines will be modeled using irregular ellipses placed at regular intervals (for each vertebra and rib pair, for instance). These ellipses, or NURBS, will be joined to generate water­tight volumes, which can be imbued with densities of various organ systems to generate precise mass estimates for extinct organisms. A particular strength of taking a digital approach to this problem, as opposed to traditional methods of building models or extrapolating from equations governing masses of living animals, is that sensitivity in results to changes within a suite of governing parameters can be undertaken simply.

Despite the completeness of Sue’s skeleton, some bones are distorted and others had loose articulations—meaning some dimensions, such as the width of the rib cage, are more uncertain than others, such as overall length. Likewise, the volume of the lungs and other low-density spaces are hard to fix, but can be modeled based on a range of proportional values from living animals, and the effect of such variation can easily be gauged with CAD models.

A “flock” of Sue models with different degrees of “fleshiness,” lung volume, and rib-cage dimensions are currently under construction and will yield a best estimate within a broader biologically plausible range of estimates. In addition to overall mass, the center of mass will be estimated, and the models will be given varying quantities of leg muscles, in order to study the effect of body size differences on estimated speed and agility. —Art Andersen

The next phase was far more complex and required quite a bit of time. We needed to make much higher resolution 3D virtual models of the individual bones. How we did this depended on the bone, especially its size and shape. Because we could not work with the original bones, as they were mounted and on display, we used a research-grade cast of each bone. These are the basic objects used by paleontologists to study dinosaurs and are incredibly accurate—taken directly from a high-quality mold of the original bone—and easier to study than the originals because they are much lighter and less fragile.

Most of these cast bones were CT-scanned; Makovicky arranged to have this done at the Loyola Medical Center in Maywood, Illinois, at the outpatient facility. It is not used on the weekends, and the staff donated their time and facilities for this project. The majority of Sue’s bones are vertebrae and chevrons, small V-shaped bones under the tail. These bones, along with some ribs and the bones of the arms, hands, legs, and feet, were scanned at the Loyola facility to provide high-resolution data. It is important to note that, most often, multiple bones were scanned simultaneously to cut down on the needed CT time.

The Chicago Police Department employed its large-format laser scanner, used for crime scenes, to digitize the popular dinosaur. (Inset) Scanning specialist Art Andersen examines one of the bones.

Not all the bone casts would fit into the Loyola medical CT scanner. The skull, hip bones, femur, and several ribs and vertebrae were CT-scanned at Ford Motor Company’s Non-Destructive Engineering Site located in Livonia, Michigan. Martin Jones, who heads up the large CT scanning facility there, used the one-meter-diameter platform to scan most of these large casts. Some of the hip bones were even too large for this CT scanner, and we contracted with a local company—Cubic-Vision of Deerfield, Illinois—to digitize these using a white-light scanner.

This variety of scanners and providers generated the data that we used to create individual models for all the real bones found as part of Sue’s skeleton. What about the few missing bones? For them, we used an approach we developed for the Smithsonian’s Triceratops project. If the missing bone is from one side of the animal—what scientists call part of its bilateral symmetry—then we used the bone from the side it was present, mirror-imaged it, and created it for the needed side. In other cases, the missing bones mostly reflect small size differences (as with some vertebrae in the spine), and for these we could simply scale the bone we didn’t have from one that we did have.

The third and final phase of the project, at least for us, is to match the higher-resolution individual cast scans with their position in the lower-resolution, full-skeleton scan made in the first phase. Working with the paleontologists at the Field Museum, we will place each individual bone scan in the exact position of its counterpart, producing a very high resolution, complete animal. This will take time, but will provide tremendous raw material for addressing questions about the biology of the animal. Now an accurate scientific study of how the dinosaur walked, ran, sat down—any question on any form of movement—can be modeled with the virtual mount. This can lead to high-quality animations of Sue moving.

As a bonus, this project also provides the potential for the museum to use the virtual models for Sue products, which can help generate much-needed funds; the options for marketing, exhibit products, and educational outreach are huge. The interest of Makovicky in answering research questions on the biology of this dinosaur not only should lead to answering those questions­—along with more general answers for tyrannosaurs, meat-eating dinosaurs, and dinosaurs in general—but should also provide ways to enhance the marketing, funding, and education missions of the owners of Sue, the Field Museum of Natural History.

For three technologists—two trained as paleontologists and one with a longtime interest in the field—it was a chance to combine two great passions and do what we all want to do: work on something we love and make a difference.

Art Andersen, president of Virtual Surfaces, has been involved in 3D scanning and digital editing for the past 15 years.

Ralph Chapman is a paleontologist and technologist, and one of the founders of the Triceratops Project, the creation of the first true digital dinosaur, at the Smithsonian’s National Museum of Natural History. He specializes in the virtualization of objects, especially those that need great care in their handling.

Linda Deck is a paleontologist and exhibition specialist, and also one of the founders of the Triceratops Project. She has comprehensive experience in incorporating 3D technology and its products into exhibits.