Skip to content 
In the field, paleontologists are sometimes shown as knowledgeable individuals stooping over the ground cleaning dust from dinosaur bones in arid, desolate settings. However, in laboratories all around the world, scientists are now uncovering the mysteries of these ancient animals using methods from medicine, chemistry, engineering, and physics.
Computer Tomography
Along with ground-breaking findings produced using these approaches during the past few years, new technologies in the field are being employed to learn more about the appearance and lives of the animals. For example, researchers are reconstructing fossils using CT (computed tomography) scanners, probing specimens using particle accelerators, and analyzing the chemical composition of samples with lasers.
Without these technical advancements, paleontologists may not have discovered that deinonychus lay blue-gray eggs, the spinosaurus was more aquatic, or that a tyrannosaurs rex utilized its huge nostrils to emit heat. These findings, along with others, could have been lost.
According to Michael Greshko, the author of a National Geographic article on the subject, paleontology in general and dinosaur research, in particular, are both in their golden years. More than 10,000 different species exist now in comparison to a group of creatures that went extinct 66 million years ago. Every day, we coexist with dinosaurs, which we refer to as birds.
In a way, it might be argued that the dinosaur era never actually came to an end. That just some of them perished while others have persisted in evolving alongside the rest of the Earth is a theory I’ll address in my following piece. Paleontologists now frequently use CT scans to rebuild missing fossil fragments without causing damage to the remaining material.
Lawrence Witmer, a paleontologist at Ohio University, used CT scans to ascertain whether or not dinosaurs expelled heat from their bodies. He discovered that the tyrannosaurus rex used its huge nasal sinuses to expel more heat using CT scanners. He said that by using CT technology to recreate missing pieces, scientists could do simulations and crash testing to learn more about how these creatures truly operated.
Witmer discovered specific groups of dinosaurs emitted heat from their bodies using the technology. He discovered the T-rex by scanning the frozen body of a Siamese crocodile that had died naturally. This allowed him to rebuild the T-internal rex’s anatomy. This technique also sheds light on dinosaurs’ growth and movement patterns 66 million years ago.
Artificial Intelligence
Paleontologists’ capacity to analyze the morphology of dinosaur remains has significantly increased thanks to the use of x-ray computed tomography. Artificial intelligence advancements mean that robots may soon be able to do the labor-intensive task of segmentation, which involves grouping like-looking portions of a picture for analysis.
Tumor detection using high-resolution medical images is among the most up-and-coming uses of AI (artificial intelligence) technology. Can scientists use the same methods to speed up the analysis of comparable dinosaur bone scans by paleontologists? In a recent publication, researchers described some of the early solutions and the ongoing obstacles.
The preserved remains’ morphology of the creatures is a major factor in how much information scientists may learn from the fossil record of dinosaurs. Cutting tiny pieces of the specimen in order to study its internal structure usually results in the sample being destroyed. Technology for high-resolution scans like CT x-ray, which essentially reconstructs inside structures in 3D (three dimensions) using computer software and radiation, brought about a shift in that.
While using CT technology aids in specimen preservation and produces a wealth of relevant data, the scans themselves have unique difficulties. The scans distinguish between different materials. Consider petrified bones as opposed to the rock that contains them.
Based on an x-ray’s radiation absorption, it may be quite difficult to tell where one thing starts and another finishes when the densities of the objects are similar. This forces researchers to manually segment objects, a time-consuming method for grouping related portions of a picture. As opposed to taking a paleontologist as long as weeks to complete, AI can separate images in a matter of minutes.
The issue remains whether a machine can classify portions of voxel (a piece of visual data representing points in 3D figures) by voxel the same as an expert would. Researchers made an effort to learn the answer using multiple kinds of deep synaptic networks, an AI model that closely resembles the human brain. More than 10,000 CT scans of three well-preserved embryonic skulls from Protoceratops, a smaller related of the more well-known species Triceratops, were used to train and test the AI systems.
The fossils were discovered in Mongolia’s Gobi Desert in the 1990s. The accuracy and processing speed demonstrated that deep neural networks could dramatically shorten the time it took to distinguish fossils from rock matrices, even if the models did not perform as well as people would have.
AI in paleontology can facilitate faster image processing in addition to establishing standards for study. The same structure may be interpreted differently by different experts, leading to distinct evolutionary history reconstructions. In some circumstances, it may be purposeful to rebuild CT pictures in a certain way. These scans may be found without significantly raising the cost by using AI segmentation.
Before that occurs, though, there is still work to be done. Even the Protoceratops test’s top model had trouble coping with other dinosaur fossils found in the same rock layers and locale. Researchers continue to train and test deep learning models using CT scans from additional fossil taxa and different preservation settings from prior digs in Mongolia. Generalization is always an issue for AI-based projects.
According to these researchers, a segmentation model for fossils from the Gobi Desert is not far away, but a more generic model requires both new algorithmic developments and more training datasets. They think deep learning will someday be able to interpret images more effectively than we can. There are currently many instances of deep learning outperforming humans in tasks like Go playing and protein 3D structure prediction.
Particle Accelerators
Some researchers transfer fossil samples to particle accelerators since some of the samples can’t fully be understood by standard CT scanners. A particle accelerator at the European Synchrotron Radiation Facility in France can see deeply within fossils and provide images that CT scans cannot produce. Magnets bend the particle stream when the electrons have completed a full cycle. The system transports electrons at speeds that are almost as fast as the speed of light.
This generates “some of the most powerful x-rays in the world,” which scientists frequently employ to examine novel substances and pharmaceuticals. Archaeopteryx (artist impression) bones were sliced through by Dennis Voeten of Sweden’s Uppsala University using the accelerator, demonstrating the species lacked the independence of flapping birds more like pheasants. This dinosaur, which is thought to have existed between 125 and 150 million years ago, represents the transition between non-avian feathered dinosaurs and contemporary birds.
I realized that you typically don’t link CT scans and particle accelerators after doing some study on them. That is the essence of modern science: it is a more collaborative endeavor than ever before. All of this is coming together in the field of paleontology to help us comprehend how dinosaurs lived.
Chemistry Advancements
Major developments in chemistry have also assisted paleontologists in determining the hue of these creatures’ feathers and the outer shells of their eggs. By using a method whereby laser light is shone over a fossilized egg, Yale University was able to see how the light scatters to disclose the underlying chemistry. The procedure was applied to the eggshells of deinonychus, highlighting two substances—protoporphyrin and biliverdin—that are also present in current bird eggs.
This discovery enables researchers to conjecture about the color of the egg, which in the instance of the deinonychus, they think was a blueish tint. The deinonychus may have utilized open-air nests, and its blue color served as a form of camouflage against the background of the sky, according to this finding.
Robot Apparatus
The finding of the spinosaurus is another fascinating development that would not have been possible without the emergence of these recently embraced technologies. It was a strange predatory dinosaur, and World War II destroyed all of its original fossils. Because there have been no fossil discoveries, there have been many concerns over what its anatomy was.
After being discovered in Egypt in the 1910s, the actual samples were destroyed by bombing some decades later, leaving only field notes, drawings, and pictures. Then, in what is now Morocco, a fresh spinosaurus skeleton was discovered a few years ago. The dinosaur’s tail, which has the appearance of a large paddle, was the skeleton fragment that was discovered.
Interestingly, they took it to a Harvard paleontologist whose colleague evaluates fish robot equipment. The scientists constructed a replica of the tail and tested it to determine how the dinosaur moved through the water. This dinosaur spent more time in the water than any other of its type, according to technology.
This is the Beginning, Not the End!
Dinosaurs are more than simply monsters from the movies or skeletal remains that show us a different world that once existed. Dinosaurs are the group to examine when examining the epic tale of how life on land copes with a changing environment. Given the current developments, it is crucial to have the baseline that dinosaurs offer.
