Image-based Bone Density Classification Using Fractal Dimensions and Histological Analysis of Implant Recipient Site

Success of dental implants is affected by the quality and density of the alveolar bone. These parameters are essential for implant stability and influence its load-bearing capacity. Their assessment is usually based on preoperative radiographs used as a tool prior to implant procedures.

The aim of the study was to compare the bone density of surgically harvested bone specimens at implant recipient sites in the maxillary and mandibular posterior region using histological analysis to the radiographic bone density using fractal dimension for reliability and determining an image based classification of bone density prior surgery.

Fifty implants were placed in the posterior region of male patients, (twenty five implants in the maxilla and twenty five in the mandible). The edentulous regions were presurgically assessed using Photo Stimulable Phosphor Plate (PSP) intra-oral radiographs and the fractal dimension box counting of region of interest was calculated at the implant recipient site. During surgery, bone core specimens were trephined, and bone densities and minerals parameters were evaluated based on histological analysis using SEM (Scanning Electron Microscopy), and atomic absorption spectrometry.

Fractal dimensions (FD) values for the same region of interest (ROI) selected on the radiographs of bone blocks and edentulous sites were different but showed a proportional variation in molar and premolar region of the maxilla and mandible. Bone density results, calculated by the ratio of bone mass (BM) to the bone volume (BV) of the bone core specimen (D=M/V), increased in the mandibular bone blocks, and decreased in the maxilla specimens. Moreover, fractal dimension values of preoperative radiographs at implant recipient sites and bone density of trephined showed a statistically similar distribution. However, no significant difference was shown in the percentage of minerals contents and mass of calcium phosphate of each bone specimen between maxilla and mandible based on scanning electron microscopy analysis. Four types of bone densities were classified according to the distribution of FD values based on preoperative radiographs and on the densities of bone cores calculations.

Radiographic estimation of bone quality calculated with fractal dimension could be a useful, non-invasive tool when using preoperative intra-oral radiographs to predict bone density at implant recipient sites with caution and limits concerning the kind of digital radiographs and size of region of interest, especially when these results were based with bone specimens harvested from implant site as an absolute reference.