Using image compression to quantify enamel complexity

Abstract

In mammals, complexity of enamel prism decussation has long been recognized as a reflection of biomechanical demands placed upon teeth. Crisscrossing prisms or bundles of prisms, known as decussation, have been demonstrated to limit crack propagation through tooth enamel, and help maintain overall functional integrity of a tooth; therefore, prism decussation tends to appear more frequently and with more complexity in animals that must break down hard foods and/or who place other high loads upon their teeth. Hunter‐Schreger bands are alternating bundles of prisms with common orientations that decussate with respect to adjacent bands. The more severe the loading regime of a tooth that a species experiences, the more complex its decussation and Hunter‐Schreger banding patterns tend to be. Species whose diets primarily create abrasive wear on teeth and who are at low risk of cracking have simple enamel whose prisms all run in parallel fashion from dentine to surface, referred to as radial enamel.In order to assess the biomechanics of teeth and microstructural adaptations to loading both within extant and fossil taxa, it would be useful to have a simple quantitative measurement that could assess overall complexity of enamel decussation. However, previous studies of enamel complexity tend to be descriptive and are often so complex in terms of their discussion of prism pathways that such studies are of limited use when comparing large numbers of teeth and/or species.The objective of this study is to solve this problem using modern, commonly available computer algorithms to find a way to quickly and easily quantify enamel complexity from histological images of teeth. Specifically, we use image compression ratio as a proxy variable for identifying tooth complexity. This ratio reflects the compressed size of a standard .jpg image as compared to an original, uncompressed .tif image (in our case generated by a digital microscopy system), and has been used successfully by prior studies to quantify complexity of tooth cusp relief.We here apply the compression ratio method to two‐dimensional photomicrographic images of tooth thin‐sections taken in circularly polarized light (CPL). CPL creates visual contrast in complex teeth by highlighting regions of differing prism orientations with different colors in the final image. We captured 90 images of teeth with varying decussation complexity in CPL, then deleted non‐enamel components of the image (dentine, artifacts, etc.) using digital image processing software. The original .tif images were then compressed to .jpg format, and the compression ratio (.jpg size / .tif size, in bytes) was analyzed statistically in comparison with subjectively assigned “complexity” categories. The compression ratio of radial enamel images was significantly different from images of decussating enamel. Moreover, we found a significant correlation between compression ratio and our complexity categories. These data suggest that image compression ratio is useful for comparing overall complexity of dental enamel in mammals.This abstract is from the Experimental Biology 2019 Meeting. There is no full text article associated with this abstract published in The FASEB Journal.