Testing the impact of two key scan parameters on the quality and repeatability of measurements from CT scan data

Abstract

Computed tomographic (CT) scanning is becoming a popular research tool across earth and life sciences. However, despite its prominence, there have not been systematic investigations into how CT scan parameters affect data quality and reproducibility. Here we conduct two sets of trials to test how exposure time, the number of x-ray radiographs averaged per view, and overall scan time affect the quality of CT scan data, assessed using signal and contrast to noise ratios and the repeatability of measurements derived from these data, in this case the calculated volume of pteropod shells. We find that contrast to noise ratio and calculated shell volume increase and the variability in shell volume measurements decrease with increasing overall scan time. However, the benefits of increased overall scan time diminish considerably at scan times of 50 minutes or more. Furthermore, as overall scan time increases, scans are at greater risk of being affected by sample movement, which can make the data unusable. By balancing exposure time and the number of x-ray radiographs averaged per view, the image quality in a 50-minute scan can be comparable to, or better than, that collected in a 75-minute scan. By selecting a 50-minute rather than a 75-minute scan, data collection can be increased by between 66 and 75%, maximizing both the quantity and quality of CT data collected.

Automatic Segmentation of Early Triassic Vertebrate Fossil CT Scans: Reducing Human Annotation Time through Deep Learning

“…In the past few decades, Computed Tomography (CT) scanning of fossil material has become the tool of choice for most palaeontologists wanting a non-destructive way to extract fragile fossils from their encasing matrix or to investigate their internal anatomy [1][2][3][4] .…”

“…With the introduction of conventional, synchrotron, and neutron micro-CT, datasets have become increasingly larger and more detailed, resulting in substantially longer post-processing times, in particular with respect to data segmentation 2,[4][5][6] With palaeontological material, where the density differences between the surrounding rock matrix and the fossil itself are usually low, manual image segmentation is often the only way to digitally extract the Regions of Interest (ROIs) from the CT slice stacks 7,8 . This process can take weeks to months to complete, thus being a considerable bottleneck in making data available for research.…”