Fig 1.
Schematic overview of dental microstructure.
Daily incremental rhythm, near-weekly rhythm, and stress-related increments are visible in histological sections of tooth enamel (Figure made by the authors using BioRender®).
Fig 2.
Example of variation in dental microstructure between two deciduous tooth sections shown at x20 magnification.
NNL = Neonatal line. E = Enamel. D = Dentine. EDJ = Enamel Dentine Junction. The NNL is generally clearly distinguishable in tooth thin sections but can vary in its manifestation. For example (A) shows the NNL as a wide and diffuse line. In contrast, (B) shows the NNL as sharp and narrow. Other Als in deciduous teeth are rarely as clear and unambiguous as the NNL. In some cases, these other ALs cannot be distinguished from regular incremental deposition lines (i.e., Retzius lines) with certainty.
Table 1.
Overview of rater profiles.
Table 2.
Definitions of AL scorings for Approach 1 and Approach 2.
Fig 3.
Total ratings of features reported by eight raters in ten histological slides for Approach 1.
The x-axis indicates the type of histological feature rated. The y-axis indicates the frequency count of the feature for a given rater. Each color represents one rater, with color choice being random. Raters who were most inclusive in their AL ratings (e.g., rater 6) had the lowest number of Retzius lines reported, meaning they interpreted the highest number of features as accentuated. Those who were the most stringent in their assessment with the lowest number of ALs (e.g., Rater 7, in orange) had the highest number of Retzius lines, as they more seldomly deemed features to be accentuated.
Table 3.
Gwet’s AC1 scores related to results of objective 1 on the reliability of raters according to level of experience and line intensity.
Table 4.
Overview of Objective 2 results on the reliability of individual vs pair ratings.
Fig 4.
Total ratings of AL types reported by six raters in five histological slides in Approach 2.
In both panels, the x-axis indicates the type of histological feature rated, and the y-axis indicates the frequency of the feature for a given rater. (A): Total ratings of AL types for individual raters before consensus, with each color presenting one rater, listed in random order. (B): Total ratings of AL types for three pairs of raters after consensus, with each color presenting one pair in random order. Those raters or pairs who were most inclusive in their AL ratings (e.g., (A) rater 6, (B), pair 1) reported the lowest number of Retzius lines, meaning they interpreted the highest number of features being accentuated compared to the other raters.
Fig 5.
Example of a section used in this study demonstrating the difference between pair ratings.
(A) Overview of a histological tooth section with the enamel in brown and dentine in dark black. The occlusal surface is facing down, with the dentine horn being exposed due to wear. The cervical enamel area shows clear ALs (red arrow). (B) Magnification of the indicated area with ALs from (A). (C, D, E) Ratings of pairs 1, 2, and 3, where each pair of two raters jointly marked the buccal enamel for AL presence and intensity. All pairs agreed on the NNL (red) and the presence of 1 MAL (off-white) as the first AL following the NNL. All pairs also marked the presence of multiple HALs in the cervical enamel (green) but had differences in how many HALs should be included in data collection.
Table 5.
Recommended workflow for the IRRISS (Improving Reliability and Reporting In Scoring of Stress-markers) approach.