In 1984, paleontologist Glenn Conroy borrowed a fossil from The American Museum of Natural History. It was about 30 million years old – not exactly young, but not old enough to be remarkable, either. It was also a hog fossil, a common one, and it seemed to Dr. Conroy unlikely that anyone would get too emotional about a ruined hog fossil.
It was the perfect fossil, in short, to be blasted with X-rays in an experiment that might go very wrong.
For most of the time that humans have been looking at fossils – and it’s a long time, dating at least back to Xenophanes’ observation that fossilized shellfish might tell a tale of when Greece was underwater – all that was known about fossils was what was apparent to the eye.
Of course, sometimes, the fossil would break, in an ambiguously happy accident: who knew what stories those exposed insides would tell? Other times, scientists just broke it on purpose.
“If you wanted to know what was inside, the only thing you could do was break it,” said Dr. Conroy, now at Washington University in St. Louis,“and curators of priceless fossils are going to be reluctant to let people do that.”
But one invention meant that curators no longer needed to let anyone ruin anything: the Computerized Tomography, or CT, scanner. Ever since Conroy’s scanned hog, digitally unwrapped from the sediment accumulations of millions of years, appeared on the cover of the journal 'Science,' CT scanners have underpinned almost every paleontological find.
CT scanning has fueled studies of the origins of flight and of dinosaurs’ preposterously long necks. It has furnished enthralling reports of new dinosaurs species. It has turned up a dinosaur heart that, whatever the sins of that dinosaur, looked all too well to scientists like true gold. It has revealed dinosaur embryos swaddled in rock, reserve teeth in dinosaurs’ dental cavities, and tail clubs primed for fending off carnivores.
All of these discoveries have underpinned paleontologists’ better than ever understanding of how these dinosaurs behaved and moved and evolved, and what this millions of years ago world looked like. Dinosaur bones are no longer just bones – with CT, they suggest how the bodies they once supported worked, how these bodies changed, and how these bodies interacted with other bodies in a complex, dynamic ecosystem. Seldom do pre-unwrapped gifts invite much excitement – what’s the fun in that? – but, to paleontologists, the gift has been in the un-wrapping.
“It’s like literally squeezing blood from a stone,” as Conroy put it.
Blood from stone
CT scanners made their debut around 1972, but did so in hospitals, for people. CT scanners use X-rays to take cross-sectional images of an object: the scanner takes a image of one layer, then moves down to take an image of the next layer, and so on. Those images are density-based, so the areas of different density show up in grey-scale shades.
Then, less then a decade after the introduction of those first, now-archaic, scanners, technology advanced enough that individual cross-sections could be digitally reconstructed into a 3D object. At about that time, it also become possible to digitally delete all the scanned material of a certain density, peeling away, say, muscles and tissues and fat to leave behind nothing but a clear, 3D model of a skeleton.
It was about that time that dinosaurs started going to the hospital.
“The first time you talk to a hospital, asking them to scan a 65 million year-old “patient,” they typically look at you as if you are from another planet,” said Robert DePalma, a paleontologist at the Palm Beach Museum of Natural History.
“I simply have to be there as it happens," says Dr. DePalma. "It’s almost like being present for a birth of a new child."
Almost three decades have passed since Conroy’s first 3D fossil model, and in that time many “children” have been born: last month, DePalma found a single T. rex tooth in a CT scan of an herbivore’s healed tailbone, a critical data point that indicated that the T. rex hunted – in this case, albeit not too well – for its meal.
And, all along, the technology has continually improved. Industrial-sized scanners have succeeded medical CT scanners in doing deeply penetrative scans of huge dinosaur bones that don’t fit into a hospital scanner. Micro CT scanners have also emerged to do the fine detail work on small fossils, taking cross-sections that are just micro-millimeters apart. Altogether, on both ends of the scale, the resolution has become sharper, the processing faster.
But most of all, those scans have accumulated.
And those vast reserves of CT data, accumulated from decades of scanning dinosaurs, have emerged as digital pots of paleontologists’ gold, the kinds that support major comparative studies alighting on new species or identifying broad evolutionary trends. The data sets have been visited and re-visited and visited again, each time with new questions about these extinct bodies worked and, by extrapolation, how this extinct world worked.
“We might scan something since we’re interested in the ear – but then someone else is interested in the nose,” says Tim Rowe, a paleontologist at the University of Texas. “You can go back to old data sets and squeeze them again.”
In 2002, Dr. Rowe had an idea: what if the CT scans taken of dinosaurs in the University of Texas’s industrial-sized scanner were put up online? What if other paleontologists could search that experimental database for the scans that would furnish their studies, saving them the time and expense of scanning a fossil for which a data set was already available?
That year, Rowe and colleagues launched the Digital Morphology library, or DigiMorph, a digital archive with QuickTime versions of the data sets. The database is an experiment in paleontological digital archiving, a prototype for a future central archive that could handle hundreds of thousands of CT scans made available to researchers for download.
But some ten years after DigiMorph went live, there is still no bona-fide archive to supplant that prototype. And DigiMorph – with some 1,200 Quicktime movies, the largest collection of biological CT scans in the world – has outgrown itself, never designed to hold so much data.
DigiMorph, and the modern technological contributions it has made to a field that studies some of the planet’s oldest once-living things, has come to represent a gap in the paleontological field: a central archive of CT scans that would function much like GenBank, the database founded in 1986 that biologists, zoologists, and geneticists consult for nucleotides. Now, paleontologists say that there is burgeoning interest in and need for creating the paleontological equivalent.
“In the same way that GenBank revolutionized genomics, what is really going to revolutionary for paleontology is a repository full of research quality data sets,” says Rowe.
At one point, the problem in getting up a central archive was a matter of protectionist foot-dragging within the paleontological community – not all researchers wanted their data put up online, says Rowe. But that resistance has begun to subside, as long as the data is not put up until after publication, he said.
“Most of the issues concern intellectual property rights, attitudes, and funding,” he says “Some researchers still guard their data closely. But the culture is changing.”
Then, for a while the problem was technological: databases couldn’t handle the huge data sets, each some 10 to 20 gigabites, which is why Digimorph archives just QuickTime movie versions of the data sets. But that too has become less relevant, he said: “the technology is on the shelf today.”
Now, the more pressing problem is that while the demand is there for a database, there just isn’t one.
Since there are still big questions about what a CT scan database would look like – what exactly would be available on the database, and would it be merely searchable or directly downloadable – paleontologists say that such a database would have to been created and coordinated by a federal organization like the National Science Foundation (NSF), which funds paleontological research.
“It’s important that we do get some sort of archive but it’s hard to get everyone to agree on what should look like,” says Amy Balanoff, a researcher at the American Museum of Natural History in New York who last month published in Nature comparative research on the evolution of the bird brain from dinosaurs. “It’s going to take something like the NSF.”
There has been some motion on the issue. In February 2013, the Office of Science and Technology Policy, and Presidential issued a mandate that federal agencies expending more than $100 million in research and development develop concrete plans for how to make the digital data that those federal funds were supporting freely available to the public. All of the proposals must include an archival solution for digital data, memo said.
“Studies on the genomic scale of anatomy and function will take massive archives and huge processing power,” says Bhart-Anjan Bhullar, a postdoctoral researcher at Harvard University, whose 2011 paper in Nature on bird evolution drew on multiple samples, some of them CT scans, to draw a multi-generation portrait of how birds got their skulls from dinosaurs.
“But imagine the depth of understanding you could gain,” he said. “The sweet spot is still ahead of us.”