AI breakthrough could revolutionize how we research dinosaur fossils

Artificial intelligence can be used to identify tumors from high-resolution medical imagery. Is it possible to use the same techniques to help paleontologists analyze similar scans of dinosaur fossils? Some of the early answers and remaining challenges were reported in a new paper.

The preserved remains of dinosaurs are the most important part of the fossil record for scientists. To study the interior structure of a specimen requires cutting thin sections. High-resolution scanning technologies like X-ray computed tomography (CT), which reconstructs internal structures in three dimensions using radiation and digital software, changed that.

The images themselves present their own challenges, while the use ofCT technology helps preserve specimen and generate very useful data. The scans differentiate various materials based on the absorption of X-ray radiation. It is difficult to determine where one object begins and another ends. Researchers have to rely on a labor-intensive process to classify similar sections of an image.

It is being put to the test.

It is possible to image segment in minutes, compared to days or weeks for a paleontologist. The question is whether a computer can do the same job as a trained professional. Deep neural networks are artificial intelligence models that mimic the human brain.

The team trained and tested the systems using three well-preserved Protoceratops skulls, a smaller relative to the more familiar Triceratops. The fossils were recovered from the Gobi Desert.

The models did not perform as well as a human, but the accuracy and processing speed showed that deep neural networks can reduce the time it takes to differentiate fossils from rock matrices.

There is need for bigger data.

Congyu Yu, lead author of the study and a PhD student at the Richard Gilder Graduate School at the American Museum of Natural History, said that using artificial intelligence in paleontology can help establish research standards. Dr. Mark A. Norell is well-known for his work on the evolutionary links between dinosaurs and birds.

Different researchers may have different interpretations of the same structure, which leads to different reconstructions of the evolutionary history. Artificial intelligence can detect frauds without increasing the cost.

There is more work to be done. The best model from the Protoceratops test was not able to perform well on other dinosaur fossils.

Yu said that researchers are continuing to train and test deep learning models on CT images from more fossil taxa and various preservation environments from previous digs.

He said that a model for fossils from the Gobi Desert is not far away, but a more generalized model is needed.

More information: Congyu Yu et al, CT Segmentation of Dinosaur Fossils by Deep Learning, Frontiers in Earth Science (2022). DOI: 10.3389/feart.2021.805271, www.frontiersin.org/articles/1 … art.2021.805271/full Citation: AI breakthrough could revolutionize how we research dinosaur fossils (2022, January 27) retrieved 27 January 2022 from https://phys.org/news/2022-01-ai-breakthrough-revolutionize-dinosaur-fossils.html This document is subject to copyright. Apart from any fair dealing for the purpose of private study or research, no part may be reproduced without the written permission. The content is provided for information purposes only.