Radiotherapy-Associated Changes in Mandibular Trabecular Microarchitecture and Radiographic Density: A Paired Fractal Dimension and Pixel Intensity Analysis

Abstract

Background: Radiotherapy for head and neck cancer can adversely affect mandibular bone through vascular injury, cellular depletion, and impaired bone remodelling. These changes may precede clinically evident complications such as osteoradionecrosis and pathological fracture. Fractal dimension (FD) and pixel intensity (PI) derived from digital panoramic radiographs may provide a simple quantitative approach for detecting alterations in trabecular architecture and radiographic bone density.

Objective: To evaluate radiotherapy-associated changes in mandibular trabecular microarchitecture and radiographic density by comparing FD and PI values on paired digital panoramic radiographs obtained before and after three-dimensional conformal radiotherapy.

Materials and Methods: This retrospective observational study included 60 patients with head and neck cancer who underwent three-dimensional conformal radiotherapy between June 2023 and June 2026. A total of 120 digital panoramic radiographs, comprising one pre-radiotherapy and one post-radiotherapy image for each patient, were evaluated. Standardized bilateral regions of interest were selected in the mandibular body. Images were analysed using ImageJ 1.54p. Trabecular complexity was assessed using FD with a box-counting method following standardized image processing, while mean grayscale PI was used as a measure of relative radiographic density. Data distribution was assessed using the Shapiro–Wilk test, and paired pre- and post-radiotherapy measurements were compared at a significance level of p < 0.05.

Results: Mean PI decreased from 131.8 ± 30.6 before radiotherapy to 112.4 ± 31.8 after radiotherapy, with a mean reduction of 19.4 units (p < 0.001). Mean FD similarly decreased from 1.41 ± 0.09 to 1.31 ± 0.10, corresponding to a mean reduction of 0.10 (p < 0.001). The concurrent reduction in PI and FD indicated a measurable decline in radiographic density together with simplification of the mandibular trabecular pattern following radiotherapy.

Conclusion: Head and neck radiotherapy was associated with significant reductions in mandibular pixel intensity and fractal dimension, suggesting deterioration in both radiographic bone density and trabecular microarchitectural complexity. Quantitative analysis of routinely acquired panoramic radiographs may offer a practical, low-cost method for monitoring mandibular bone changes after radiotherapy. Further studies incorporating radiation dose, treatment field, and longer follow-up are required before FD and PI can be established as predictive markers of radiation-related mandibular complications.