The evaluation of implant stability is a crucial task for dentists. The long-term stability of a dental implant is dependent on the quality of bone tissue surrounding the implant. In this context, difficulties are mostly given to the complexity of newly formed bone tissue (a complex, anisotropic, porous-viscoelastic medium in constant remodeling) but also to the boundary conditions at the bone-implant interphase. In particular, multiple parameters affect primary and secondary stability, as bone quality, bone density or amount of bone in contact with the implant.

In this framework, ultrasound based techniques have been proven to be effective in the qualitative and quantitative evaluation of primary and secondary stability of dental implants. The ultrasonic response of the implant depends on the microstructure, mechanical properties and geometry of the bone-implant system. Major questions are: how each of these properties leaves a specific signature on the signal, and how the information can be extracted. One must conceive advanced mechanical models describing the target sample associated with relevant signal processing techniques.

The purpose of this work is to process ultrasonic signals obtained from dental implants signals in different environments (i.e. for different sets of parameters) using advanced signal processing techniques, based on multifractal analysis.

Implants with different stabilities, obtained by suitably unscrewing the dental implant, are simulated using a finite element software. The advantage of numerical simulations with respect to experiments is that a sensitivity analysis with respect to parameters such bone density and stiffness can be performed in a controlled manner.

Both experimental (for wet bone and dry bone) and numerical (a parametric study varying the frequency and the trabecular bone density has been considered) results are compared, and then analyzed by signal processing tools based on multifractal methods. In this context, the aim of multifractal analysis is to describe both signal scale properties and its local singularities distribution. So, it allows to access to the internal structure of the signal. The aim is to extract relevant parameters able to characterize the composition and/or the micro-structural properties of bone-implant inter-phase (self-similarity properties). By looking at the mean values of the structure functions to the level of unscrewing (of the implant), a correlation can be observed.