Trabecular Bone Fractal Dimension Adjacent to Unilateral Vertically Impacted Mandibular Third Molars: A Split-Mouth Study

Abstract

Background/Objective: Bone is a mechanically responsive tissue whose trabecular microarchitecture adapts to local functional loading. Whether the presence of an unerupted, unilateral, vertically impacted mandibular third molar is associated with detectable local physiological remodeling of the adjacent trabecular bone has not been well characterized. This study compared the trabecular bone fractal dimension (FD) adjacent to unilateral vertically impacted mandibular third molars with the corresponding contralateral mandibular region, using standardized regions of interest (ROIs) on panoramic radiographs from an open-access dataset. Methods: Fifty panoramic radiographs meeting predefined inclusion criteria were retrieved from the publicly available m-TM (Mandibular Third Molar) dataset in a split-mouth design. Two standardized 64 × 64-pixel ROIs were placed adjacent to the impacted third molar and at the corresponding contralateral region. Image preprocessing and box-counting fractal analysis were performed in ImageJ (version 1.54g; National Institutes of Health, Bethesda, MD, USA) with the FracLac plugin for ImageJ (version 2015Sep09; A. Karperien, Charles Sturt University, Australia). Normality was assessed with the Shapiro–Wilk test; paired comparisons used a paired-samples t-test. Results: Mean FD adjacent to the impacted side was 1.449 ± 0.144, compared with 1.447 ± 0.136 on the contralateral control side. Paired differences were normally distributed (W = 0.963, p = 0.113). No statistically significant difference was found between sides (mean difference = 0.0019, 95% CI: −0.021 to 0.025; t (49) = 0.168, p = 0.867). The narrow confidence interval argues against a measurable undetected difference, although no established clinical-meaningfulness threshold exists for fractal dimension and formal equivalence testing was not performed; as a secondary, standardized effect-size measure, Cohen’s d was negligible but imprecisely estimated (d = 0.024, 95% CI: −0.254 to 0.301). Conclusions: Within the resolution of two-dimensional panoramic fractal analysis, no detectable difference in fractal dimension was observed between the region adjacent to a unilateral vertically impacted mandibular third molar and the contralateral region. This finding is consistent with the possibility that an impacted, functionally unloaded tooth provides insufficient mechanical stimulus to elicit a radiographically detectable change by this method, an interpretation discussed in relation to Wolff’s law and mechanotransduction-based bone physiology but not directly tested in this study.