Identification of root filling interfaces by microscopy and tomography methods

Abstract

Aim To assess differences in observed cross-sectional areas of root canals and filling materials, as imaged by three microscopy and two tomography methods. Methodology Six roots filled with laterally compacted Gutta-percha and AH26 were scanned with phase-contrast enhanced microtomography in a synchrotron facility. Reconstructed virtual slices were compared with sections of both wet and acrylic-embedded roots, evaluated also by light and electron microscopy (EM) and laboratory-based microtomography (μCT). The different contrasts of Gutta-percha, voids, sealer and root dentine were identified and correlated. Inner canal border, outer Gutta-percha rim and the external margin of a void were manually delineated, and the enclosed areas were repeatedly measured by three observers. Interobserver and interimaging method differences were tested by 2-way anova with Bonferroni adjustments (P<0.05). Percentages of Gutta-percha-filled canal areas (PGP) were determined. Results Phase-contrast enhanced microtomography revealed internal interfaces and detailed 3D volumes of accentuated voids as well as micrometre-sized particles and gaps within the treated roots. Overestimates in the cross-sectional areas were obtained by light microscopy, whereas underestimates were obtained by μCT and EM. Differences exceeded 40%; however, PGP values by all methods were within 5% for the same slice. Differences between observers were sometimes significant, but they were not method related (<3%). Conclusions Phase-contrast enhanced microtomography is a powerful non-destructive ex vivo investigation method for studying the interfaces within root canals and filling materials at a micrometre resolution. The method does not require damage-prone sectioning/polishing during sample preparation procedures. Caution should be used when quantifying the extent of Gutta-percha in root fillings by measurements using μCT, light and EM. © 2010 International Endodontic Journal.