Clinical studies have found high success rates for dental implant treatment, but the failure rate of dental implants remains high in poor quality bone. Usually the bone quality and physical properties are better for the mandible than for the maxilla, and superior for the anterior region than for the posterior region. This has resulted in research findings that the survival rate is higher for dental implants in the mandible than in the maxilla, particularly in the anterior region.

Bone quality is clinically classified into types 1–4 according to Lekholm and Zard based on the amounts of cortical bone versus cancellous bone evident on radiograph film. There are reports of only 3% of bone being lost in implants placed in bone of type 1, 2, or 3, whereas the failure rate was 35% in type 4 bone, which has a thin cortical shell and constitutes softer cancellous bone. The implant stability is lower in type 4 bone than in bone of other qualities. Orthopaedic and dental implant research have indicated that the stiffness of cancellous bone significantly influences the holding strength of an implant. In the clinical situation, the Lekholm and Zarb classification only provides a rough assessment of the quality and quantity of jaw bone, and hence precisely how bone quality and quantity affect implant stability is still unclear.

Micro-level computed tomography (micro-CT) is now widely used for observing and analyzing the internal structure of hard tissue due to its advantages of rapidity, reproducibility, and nondestructiveness. Many studies have used micro-CT to obtain high-resolution images and assess the trabecular structure of human bone quantitatively in three dimensions (3D). Micro-CT has also been increasingly employed in research related to dental implants to evaluate the peri-implant bone, and has been validated using histomorphometric results. Micro-CT also allows for accurate 3D measurements of bone-to-implant contact (BIC). Only a few studies have assessed the 3D BIC, and no study has investigated its correlation with initial implant stability.

Resonance frequency analysis (RFA), which was introduced by Meredith et al., can measure the initial stability of a dental implant nondestructively. This is performed using the Osstell instrument (Osstell, Göteborg, Sweden), and has become a common method for determining the initial stability of a dental implant in both in vitro and clinical studies. Many studies have used histological methods to measure two-dimensional (2D) BIC and investigate its relationship to implant stability as measured by RFA. The 2D BIC obtained from a histological section may not accurately represent the entire 3D interface between implant and bone, and no previous study has used micro-CT to evaluate the relation between the 3D BIC and RFA values (usually evaluated as the implant stability quotient (ISQ)).

The present study used high-resolution micro-CT to determine how initial implant stability, as measured by RFA, is related to the 3D BIC ratio (BIC%) and bone stiffness (elastic modulus). This study also examined differences in implant stabilities between cancellous bone with and without a cellular structure.

Bone sample preparation

Sawbone models of one solid (elastic modulus of 123 MPa) and four cellular (elastic moduli of 137, 47.5, 23, and 12.4 MPa) polyurethane foam blocks (models 1522-02, 1522-09, 1522-10, 1522-11, and 1522-12, Pacific Research Laboratory, Vashon Island, WA, USA) were used to mimic the cancellous bone of the jaw. The trabecular bone blocks were attached to 2 mm thick commercially available synthetic cortical shell (model 3401-01, Pacific Research Laboratories) with an elastic modulus of 16.7 GPa to represent the experimental bone models of cancellous bone with varying densities ( Fig. 1 ). All of the experimental bone models were rectangular with dimensions of 38 mm × 20 mm × 42 mm. A commercial dental implant (4 mm in diameter and 12 mm long; ATLAS Implant System, Cowell Medi, Busan, South Korea) ( Fig. 2 ) was used as the target dental implant in this study. Five implant specimens were prepared for each of the four groups (comprising artificial foam bones with four elastic moduli) for the measurements of implant stability and BIC%.

Models of cancellous bone used in the experiments: (a) 123 MPa solid-rigid polyurethane-foam block; and (b) 137 MPa, (c) 47.5 MPa, (d) 23 MPa and (e) 12.4 MPa cellular blocks.

ISQ values of implants were measured by using the wireless Osstell ISQ system.

3D BIC analysis by micro-CT

For calculating the 3D BIC%, micro-CT (SkyScan-1076, Skyscan, Aartselaar, Belgium) was used to obtain 3D information from each specimen with an implant at a resolution of 17.2 μm × 17.2 μm × 17.2 μm. Scans were performed on approximately 800 slices, with 1024 × 1024 pixels for each slice. The micro-CT images of each specimen with an implant were imported into professional medical imaging software (Mimics 13.0, Materialise, Leuven, Belgium) ( Fig. 3 a ) and were segmented by using different thresholds for implant and bone to construct a 3D computer model ( Fig. 3 b). The whole exterior surface of the implant inside the specimen and the BIC area were determined, with the 3D BIC% calculated as the area of the BIC region divided by the exterior surface of the implant.

(a) Micro-CT image in middle section of the implant. (b) 3D computer model of the foam bone (including cortical bone and cellular cancellous bone) with an implant.