ORIGINAL ARTICLE Evaluation of trabecular bone healing by fractal analysis and digital subtraction radiography on digitized panoramic radiographs: a preliminary study H. Koca • S. Ergu ¨n • P. Gu ¨neri • H. Boyacıoglu Received: 11 September 2009 / Accepted: 10 November 2009 / Published online: 18 December 2009 Ó Japanese Society for Oral and Maxillofacial Radiology and Springer 2009 Abstract Objectives We have evaluated radiographic changes in mandibular bone texture before and after surgical therapies using fractal analysis and digital subtraction radiography (DSR). Materials and methods Preoperative and postoperative panoramic radiographs of 10 patients were acquired and converted into digital format. The procedures for calcu- lating the fractal dimension (FD) were performed using Image J 1.38x software, and the mean gray values (MGVs) for digital subtraction were implemented using newly developed software (RAIN). Data were analyzed statisti- cally with the paired t-test and Pearson’s correlation test. Results The differences between the FD values (p = 0.0019) and between the MGVs (p = 0.0181) of the preoperative and postoperative images were statistically significant; the correlation did not reach statistical signifi- cance, however (p [ 0.05). Conclusion FD analyses and DSR application are pro- posed for quantitative determination of bone changes. Further studies are planned to provide a more detailed evaluation of these methods. Keywords Bone tissue Á Fractals Á Subtraction radiography Á Digital image processing Introduction Visual comparison of paired radiographs taken at different times creates confusion when identifying the radiographic features that are solely related with the progression or regression of a particular disease, mainly because of the structural noise that results from anatomical details pro- jected over the area of radiographic change [1, 2]. Addi- tionally, the variations involved with conventional radiographic exposure, object/film/X-ray tube alignment, and choice of region of interest (ROI) result in higher error rates [3, 4]. Innovative digital dental image applications enable more precise estimation of disease extent and prognosis [5–7]. Of the several methods that have been developed to estimate alveolar bone change from dental radiographs, fractal analysis and digital subtraction radiography (DSR) seem to be the most economical and easily available methods [8–12]. DSR emphasizes the differences between pairs of radiographs by removing structural noise. The resulting image displays the areas of bone change as either dark or light shades of gray [13, 14]. During DSR, it is critical to eliminate differences in image alignment and contrast to provide a perfect subtracted image, and this is usually regarded as a limiting factor of this method in clinical practice [9]. With the introduction of geometric recon- struction algorithms [10, 15], a posteriori registration H. Koca Department of Oral and Maxillofacial Surgery, Ege University School of Dentistry, Bornova, Izmir 35100, Turkey S. Ergu ¨n Á P. Gu ¨neri (&) Department of Oral Diagnosis and Radiology, Ege University School of Dentistry, Bornova, Izmir 35100, Turkey e-mail: peleen_2000@yahoo.com H. Boyacıoglu Department of Statistics, Ege University Faculty of Science, Bornova, Izmir 35100, Turkey 123 Oral Radiol (2010) 26:1–8 DOI 10.1007/s11282-009-0029-6