An international team led by CENIEH has analyzed nearly a thousand baboon skull scans and found that around 20% contained hidden errors, highlighting the need for stricter quality controls
An international team led by Mario Modesto-Mata, a scientist at the Centro Nacional de Investigación sobre la Evolución Humana (CENIEH), has just published a study in the journal PeerJ, warning about the unexpected challenges of working with large 3D digital image repositories. After analyzing nearly a thousand baboon skull scans available on the open platform MorphoSource, the study’s authors discovered that about 20% contained hidden errors that could severely distort scientific results.
“We realized the problem when importing the images into the 3D Slicer program for analysis. There, the files triggered a warning about inconsistencies in the scanner slice sizes, which could alter skull measurements by up to 35%. If overlooked, these failures could have compromised years of work and conclusions about primate evolution and genetics," says Mario Modesto-Mata, member of the Tied2Teeth Group and affiliated with the International University of La Rioja (UNIR).
To confirm the accuracy of the data, the authors compared the digital measurements with those of real physical skulls, using traditional tools such as calipers. Following eight months of intensive work and more than 250 hours of effort, they successfully preserved a unique dataset, ensuring its usefulness for future research.
The study not only warns that digital errors are more common than they seem, but also sends a clear message: open data presents great opportunities, but it requires stricter quality controls and self-criticism from researchers when using it.
“It's tempting to ignore an error message and move on, but doing so can come at a very high cost. Our study shows that investing time in reviewing the data ensures reliable results and avoids misleading conclusions. That's why we suggest that storage platforms incorporate automatic checks, and that funding agencies recognize the extra effort involved in preparing and validating this type of information," says Modesto-Mata.
This study, funded by the European Research Council (ERC) as part of the Tied2Teeth project led by Leslea J. Hlusko, serves as a reminder to both the scientific community and the general public: digital science opens new doors, but it also demands patience, rigor, and a commitment to quality.
