Adult age-at-death estimation using the pulp/tooth area ratio (PAR) method: a South African validation study using periapical radiography and stereomicroscopy

Kotze, Daniël

Adult age-at-death estimation using the pulp/tooth area ratio (PAR) method: a South African validation study using periapical radiography and stereomicroscopy

Thesis / Dissertation

2024

Publisher

University of Cape Town

Department

Department of Human Biology

Faculty

Faculty of Health Sciences

Abstract

Conventional skeletal age-at-death estimation methods employed for South African adults are deficient, facing several limitations. Dental ageing techniques remain under-researched for the population. In this study, I aimed to evaluate the applicability of the pulp/tooth area ratio (PAR) ageing method when applied to maxillary canines from a South African cadaveric sample. I explored the utility of under-researched methodological approaches for the method: mesiodistal radiographs, stereomicroscopic tooth section images and PARs excluding the enamel area. To maximise sample size, only maxillary canines were considered, which demonstrate several research advantages (e.g. relatively long functional survival and high accuracy rates). Following extraction, 52 adult teeth were radiographed, sectioned and analysed using stereomicroscopy. Labiolingual and mesiodistal radiographs and labiolingual stereomicroscopic tooth section images were captured and analysed using ImageJ to obtain PARs. Age estimation linear regression models were developed for ratios obtained from each image type, including and excluding the enamel area. Models were compared for performance and accuracy using best subsets regression and cross-validation analyses. The accuracy of international standards was also assessed using the study sample. The PAR method showed promising results. It demonstrated excellent observer reproducibility and sex-independence. Sample-specific models derived from tooth section images had the best performance and accuracy, obtaining cross-validation mean absolute errors (MAEs) and standard error of the estimates (SEEs) of approximately 7 - 8 years and 10 - 11 years, respectively. Labiolingual radiographic models (MAEs = 10 years; SEEs = 12 years) showed better performance and accuracy than mesiodistal radiographic models (MAEs = 12 years; SEEs = 16 2 years). Models excluding the enamel area performed better, producing slightly better cross validation error values. Standard PAR models performed relatively poorly and were inappropriate for the study sample. This research suggests that the PAR method is a suitable technique to complement and inform standard adult skeletal age analyses in South Africa. The labiolingual radiographic approach is most practical as it is relatively accurate, minimally invasive and efficient. This study contributes knowledge to the limited pool of data on dental ageing techniques for South African adults. It informs practitioners of ideal/optimal methodological approaches for the PAR method using maxillary canines.

Keywords

Medicine

Reference:

Collections

Masters

Full item page