Fractal analysis of dental radiographs to detect periodontitis-induced trabecular changes

Abstract

Background and Objective: The aim was to evaluate fractal analysis as a tool to quantitatively measure the impact of periodontal disease on surrounding bone. The diagnosis of periodontitis is based on information obtained from clinical and radiographic examinations. The current standard use of dental radiographs is visual inspection, often with no quantitative analysis. Fractal analysis can be used to examine trabecular bone patterns among periodontal patients. Material and Methods: Patients (n = 108) from the University of Southern California School of Dentistry were classified into three groups: healthy, moderate and severe periodontitis. A region of interest was selected from periapical radiographs. Image processing was applied to correct for lighting irregularity, and the box-counting method was used to calculate a fractal dimension. ANOVA and ANCOVA were used to measure fractal dimension differences between all groups. Results: According to the statistical tests, significant differences in average fractal dimensions were measured between healthy and moderate periodontitis groups (p < 0.01) and between healthy and severe periodontitis groups (p < 0.001). Higher fractal dimensions were measured in healthy periodontal patients. Conclusion: Fractal analysis evidenced significant differences between patients affected and not affected by periodontitis. The box-counting method quantitatively describes the severity of bone disease and can be used to improve current diagnostic techniques. © 2008 Blackwell Munksgaard.