Bone density and proximal support effects on dental implant stability - Finite element analysis and in vitro experiments

OBJECTIVES

The application of dental implants presents the occurrence of implant failures associated with bone proximal support. This study aims to assess implant behavior, in particular implant stability and strain distribution in the bone at different bone densities, and the effect of proximal bone support.

MATERIAL AND METHODS

Three bone densities (D20, D15, and D10) were considered in the experimental in vitro study, represented by solid rigid polyurethane foam and two conditions of bone support in the proximal region. A finite element model was developed and validated experimentally and a Branemark model at a 3:1 scale was implanted in the experiments; the model was loaded and extracted.

RESULTS

The results of the experimental models validate the finite element models with a correlation R2 equal to 0.899 and NMSE of 7%. The implant extraction tests for the effect of bone properties in the maximum load were 2832 N for D20 and 792 N for D10. The effect of proximal bone support changes the implant stability was observed experimentally; at 1 mm less bone support decreases by 20% of stability and at 2 mm by 58% for D15 density.

CONCLUSIONS

Bone properties and bone quantity are important for the initial stability of the implant. A bone volume fraction of less than 24 g/cm3 exhibits poor behavior and is not indicated for implantation. Proximal bone support reduces the primary stability of the implant and the effect is critical in lower bone density.

The effect of osteochondral lesion size and ankle joint position on cartilage behavior - numerical and in vitro experimental results

Osteochondral lesion of the talus is defined as damage in the cartilage that covers the talus bone, compromising the integrity of the joint in the long term. Due to the low incidence of this pathology, there are few studies to understand the importance of lesion size and position in cartilage strains. The purpose of this study is then to analyze the influence of the lesion size in joint behavior. A 3D virtual and in vitro model of a patient's injured ankle joint was developed. The models were built using CT scan and MRI images, to obtain the CAD models of intact and with 10 mm lesion size for 3D print models using additive manufacturing. The physical model was tested with 685N applied vertically to determine experimentally the principal strains and contact pressures in the cartilage. Five finite element models were developed with lesion dimensions (5 to 20 mm) and with 3 ankle joint positions. The numerical and experimental results were correlated with an R² = 0.86 justified by the complexity of the model geometry. The maximum principal strain was 2566µε in the plantar flexion position without lesion. The experimental contact area between cartilages increased by 1.2% in the 10 mm lesion size for 431 mm². The maximum stress in the cartilage was observed for a 20 mm lesion size with 2.5 MPa. The 5 and 10 mm sizes present similar results; the 15 mm lesion size presents a stress increase of 13% comparatively with 10 mm. Plantar flexion seems to be the most critical configuration; stress increases with an increase of lesion size around the cartilage.

Influences of geometrical and mechanical properties of bone tissues in mandible behaviour - experimental and numerical predictions

The properties and geometry of bone in the mandible play a key role in mandible behaviour during a person's lifetime, and attention needs to be paid to the influence of bone properties. We analysed the effect of bone geometry, size and bone properties in mandible behaviour, experimenting on cadaveric mandibles and FE models. The study was developed using the geometry of a cadaveric mandible without teeth. Three models of cadaveric condyles were experimentally tested with instrumented with four rosettes, and a condyle reaction of 300 N. Four finite element models were considered to validate the experiments and analyse mandible behaviour. One numeric model was simulated with 10 muscles in a quasi-static condition. The experimental results present different condyle stiffness's, of 448, 215 and 254 N/mm. The values presented in the rosettes are influenced by bone geometry and bone thickness; maximum value was -600 με in rosette #4, and the maximum strain difference between mandibles was 111%. The numerical results show that bone density decreases and strain distribution increases in the thinner mandible regions. Nevertheless, the global behaviour of the structure remains similar, but presents different strain magnitudes. The study shows the need to take into account bone characteristics and their evolutions in order to improve implant design and fixation throughout the patient life. The change in bone stiffness promotes a change in maximum strain distribution with same global behaviour.

Experimental and numerical predictions of Biomet(®) alloplastic implant in a cadaveric mandibular ramus

The purpose of this study was to evaluate experimentally the behaviors of an intact and an implanted cadaveric ramus, to compare and analyze load mechanism transfers between two validated finite element models. The intact, clean cadaveric ramus was instrumented with four rosettes and loaded with the temporal reaction load. Next, the Biomet microfixation implant was fixed to the same cadaveric mandibular ramus after resection. The mandibular ramus was reconstructed from computed tomographic images, and two finite element models were developed. The experimental results for the mandibular ramus present a linear behavior of up to 300 N load in the condyle, with the Biomet implant influencing strain distribution; the maximum influence was near the implant (rosette #4) and approximately 59%. The experimental and numerical results present a good correlation, with the best correlation in the intact ramus condition, where R(2) reaches 0.935 and the slope of the regression line is 1.045. The numerical results show that screw #1 is the most critical, with maximum principal strains in the bone around 21,000 με, indicating possible bone fatigue and fracture. The experimental results show that the Biomet temporomandibular joint mandibular ramus implant changes the load transfer in the ramus, compared to the intact ramus, with its strain-shielding effect. The numerical results demonstrate that only three screws are important for the Biomet TMJ fixation. These results indicate that including two proximal screws should reduce stresses in the first screws and strains in the bone.

Influences of implant condyle geometry on bone and screw strains in a temporomandibular implant

A 3D finite element model of an in vitro implanted mandible was analysed. The load point was placed on the condyle in three positions (inside the mouth, centred and outside) to simulate different contact points between the mandible condyle and the temporal bone. The strain fields in the condyle were assessed and detailed around the surgical screws. The temporomandibular implant studied here was modelled on a commercial device that uses four screws to fix it in vivo in a very similar position. The boundary conditions of the numerical model simulated a load on the incisors with a 15 mm mouth aperture. The same contact loads were applied to the two condyles. Numerical results were successfully obtained for the three different contact points: the inside contact produced lower strains on the condyle. The first screw created a critical strain distribution in the bone, just under the screw. The study shows that centred and inside contact induces lower strain distributions. This suggests that spherical condyle geometry should be applied in order to reduce the strains in fixation. As the top screw was observed to play the most critical role, the third screw is in fact unnecessary, since the lower strain distribution suggests that it will be loosened.