Department of Anthropology, Albert-Ludwigs-University of Freiburg - Medical Center
Saturday Afternoon, 301D
Tooth cementum annulation (TCA) is amongst the most precise anthropological age determination methods to date. In most TCA studies the incremental lines in tooth cementum are counted manually by various observers. This can lead to large inter- and intraobserver errors. In order to overcome observer bias and to receive statistically significant results, Czermak et al. developed in 2006 a semi-automated counting software.
The study at hand evaluates this software by applying it to 306 images of 65 soil-exposed teeth of 65 individuals with known age from the Spital field in Basel, Switzerland (19th century). Each image was counted with the software and an average age from all images of one individual (3-10 images per individual) was calculated. This process was repeated to test the software reliability.
The semi-automated software results indicate a clear trend of age underestimation. Both counts show stable results, with an average age deviation of -7,43 years (absolute 8,98 years) for the first count and an average age deviation of -7,78 years (absolute 9,09 years) for the second count.
Compared to the standard TCA method count results of 3 observers that range from an average absolute age deviation of 3,59 years (trained observer I) to 14,54 years (untrained student observer III), the actual software cannot reproduce the good results of a trained observer. However, the first evaluation of this software gives hints on how to improve it to make it a useful tool for anthropologists, especially for untrained observers.