Skip to main content

Patient-specific predictions for bone treatments

Final Report Summary - CAD-BONE (Patient-specific predictions for bone treatments)

CAD-BONE: Patient-specific predictions for bone treatments

The main objective of CAD-BONE was to perform a multidisciplinary research that resulted in a technology that radically improves the development of patient-specific computer models for the modelling of bone adaptation/healing after prosthesis implantation in different bone treatments for musculoskeletal applications. Therefore, the final aim of CAD-BONE was to demonstrate the feasibility of integrating patient specific modelling, musculoskeletal loading, and simulation of adaptive bone remodelling to simulate functional outcome of patient treatments.

This multidisciplinary approach has allowed the achievement of the following complementary specific objectives, which go substantially beyond current state-of the-art:
Objective 1: Development of patient-specific FE models: Integration of different computer technologies with a predictive role.
Objective 2: Quantification and verification of CAD-BONE multidisciplinar computer approach of bone remodelling/healing models.
Objective 3: Predictive computer simulations to estimate bone response after surgical intervention.

CAD-BONE combines image processing, musculoskeletal modelling tools, finite element analyses and bone remodelling/healing algorithms to provide an understanding of the individual functional outcome of patient treatments from standard clinical radiographs. Although already developed and broadly in use, these computer technologies and methods are combined and integrated to create a computer tool with a predictive purpose. Different scientific and technological challenges have been faced to successfully achieve this objective.

Therefore, CAD-BONE has created a preliminary tool for the prediction of bone morphological changes resulting from changes in mechanical loading conditions. More information can be found in the project web page (

Since the beginning of the project, we have created a specific model of the femur where the identification of muscle attachment regions and forces has been done (by gait analysis). We have also validated these loads through a methodology that combines gait analysis with bone remodelling algorithms. We have clearly demonstrated that loads estimated from gait analysis are able to predict the bone density distribution, which resulted very similar to the one determined by the CT data, clearly showing the potential of this methodology. The methodology was successfully applied to the prediction of tibia loading conditions. The results obtained were validated with gait analysis data.

Materialise has implemented in Mimics different suggestions that clearly improved their software. The main incorporations have been: the automatic mesh generation and the FE pre-processor for connecting their software with ABAQUS and MARC.

CAD-BONE has analyzed and implemented different phenomenological-based bone remodelling models (Nijmegen’s model, Stanford’s model and anisotropic bone remodeling). We have observed several limitations on these models. There is a strong dependency of the final bone remodelling solution on the initial conditions fixed in the simulations. Therefore, different alternatives have been considered in order to solve the problem.

We have also developed a phenomenological bone healing model able to simulate the healing of complex bone fractures (oblique, comminuted, with several fragments). The model has been validated with several examples of application and with a clinical case where the final outcome of the fracture is known. The results have clearly demonstrated its potential for modeling the fracture healing.

One of the main tasks to be solved by the project was that bone remodelling/healing computer models were very slow, labour-intensive and costly processes. We investigated different numerical strategies to cut this time of analysis. Two techniques have been developed and applied to accelerate the bone remodeling simulations (Artificial Neural Networks-ANN and Extrapolation Methods). We have achieved to accelerate bone remodeling simulations with a significant impact with these two methods.

Finally, a parametric finite element model (in geometry and density) of the femur was developed and a sensitivity analysis of the proximal femur parameters was performed in order to check the influence of every model parameter on the fracture risk prediction.

All this work has been published in different journals (5 papers, 2 under review) and in relevant international conferences and scientific forums. Additionally, 3 PhD theses were developed within the project framework and one is in its latest state.

The most valuable impact of CAD-BONE has been that its three partners, UNIZAR, KULEUVEN and MATERIALISE, have become more competitive after the project completion as they have improved many of their current abilities. Based on their previous respective research and development collaborations, the knowledge gained and transferred throughout CAD-BONE and the enhanced synergies and complementarities, the three partners have been well prepared to face future joint research collaborations.

From a technological point of view CAD-BONE have demonstrated the feasibility of integrating patient- specific modelling, musculoskeletal loading and bone remodelling/healing ability in order to predict the functional outcome of some prosthetic treatments. From a scientific point of view, CAD-BONE have developed phenomenological-based bone remodelling and healing numerical models, improving the current state-of-the-art regarding the role of mechanical factors in such biological processes. From a socio-economic point of view the results of CAD-BONE have had an impact aligned with the Europe 2020 strategy, specifically with the idea of smart growth, helping to improve EU’s performance in research and innovation on such a major challenge for our society as health, by 1) Creating new products and future services and 2) Strengthening the link between research and commercialization through effective knowledge transfer and close collaboration between the RTD and the commercial partners.

The scientific and technological reach of CAD-BONE falls within one of the four market categories identified by the eHealth Lead Market Initiative: Clinical Information Systems, as it deals with the development of specialized tools for health professionals in the fields of medical imaging systems and computer-assisted surgery training and planning systems. By integrating different technological solutions, CAD-BONE has help tackling the fragmentation of the market due to poor communication between existing components, easing the take-up of new products by the users.