Elsevier

Computers in Biology and Medicine

Volume 52, 1 September 2014, Pages 8-17
Computers in Biology and Medicine

Experimental validation of 3D printed patient-specific implants using digital image correlation and finite element analysis

https://doi.org/10.1016/j.compbiomed.2014.06.002Get rights and content

Highlights

  • We design patient-specific bone replacement shapes using topology optimization.

  • We perform mechanical testing on a 3D printed bone replacement by applying masticatory forces.

  • We validate the strains of the tests with digital image correlation and finite element analysis.

  • We show that the topology optimized implants provide adequate load-transfer mechanism.

  • This technique has the potential to alleviate the uncertainty for bone replacement designs.

Abstract

With the dawn of 3D printing technology, patient-specific implant designs are set to have a paradigm shift. A topology optimization method in designing patient-specific craniofacial implants has been developed to ensure adequate load transfer mechanism and restore the form and function of the mid-face. Patient-specific finite element models are used to design these implants and to validate whether they are viable for physiological loading such as mastication. Validation of these topology optimized finite element models using mechanical testing is a critical step. Instead of inserting the implants into a cadaver or patient, we embed the implants into the computer-aided skull model of a patient and, fuse them together to 3D print the complete skull model with the implant. Masticatory forces are applied in the molar region to simulate chewing and measure the stress–strain trajectory. Until recently, strain gages have been used to measure strains for validation. Digital Image Correlation (DIC) method is a relatively new technique for full-field strain measurement which provides a continuous deformation field data. The main objective of this study is to validate the finite element model of patient-specific craniofacial implants against the strain data from the DIC obtained during the mastication simulation and show that the optimized shapes provide adequate load-transfer mechanism. Patient-specific models are obtained from CT scans. The principal maximum and minimum strains are compared. The computational and experimental approach to designing patient-specific implants proved to be a viable technique for mid-face craniofacial reconstruction.

Introduction

Craniofacial segmental defects that are caused due to blast injury or tumor ablation require reconstructive procedures involving large bone replacement implants. Such defects remain a challenging problem for reconstructive surgeons as it is difficult to create a complicated 3D structure that satisfies significant functional and aesthetic roles of the midface. Proper reconstruction and replacement are needed not only to obliterate the defect, but also to address the issues of swallowing, orbital function, vision, mastication, speech and restoration of facial contour and self-image [1]. These defects are patient-specific and the implants should depend on the loads and the dimensions of the missing bone defect. Large segmental bone defects of the midface typically require bone grafts for a successful outcome and, advances in microvascular free flaps have greatly increased reconstructive options. Virtual surgical planning, stereolithography modeling, and prefabricated osteotomy bone cutting templates have made surgical planning of osteotomies easier [2], [3], [4]. However, for most of these cases, no structural analysis or design principles are used which often results in unfavorable circumstances such as fracture or high stress concentrations. Advances in 3D printing technology have opened a new dawn in patient-specific implant design. Recently, a bioresorbable airway splint has been successfully created and implanted in an infant using a 3D printer [5]. This work demonstrated innovative creation of implants for patient-specific anatomical condition using a multi-disciplinary approach involving high-resolution imaging, computer aided design and biomaterial based 3D printing. For a large segmental defect, finding the correct and optimized shape of the bone replacement still remains a question. Reconstruction is typically done heuristically without adequate surgical preplanning and structural analysis. To address this problem, a technique named topology optimization has been introduced into designing bone replacement in our previous work [6], [7]. Topology optimization is a method of finding a structure with optimal load paths to transfer a number of loads to defined supports. This method is widely used in designing components for the automobile industry and lately in designing aircraft components. For example, it has been successfully applied to design the main wing box rib of Airbus A380 [8].

Surgical outcome can be enhanced by taking an interdisciplinary approach. Custom tissue fabrication can be considered an option to replace the region of defect ideally [9]. A potential strategy to achieve this objective for craniofacial reconstruction is to apply bone formation technique to patient specific solutions obtained from computer aided design (CAD) with computational methods [7]. The solution is optimized for functional restoration when combined with soft tissue repair and appropriate prosthetics even though the design may deviate from the original human bone geometry. The objective function of topology optimization algorithm is to minimize compliance within the structure. Although only mechanical loadings were considered, it was found that topology optimization gives a shape similar to human natural skeleton without a defect when applied to cross-section on a vertical plane through first molar [7]. Understanding how functional physiological loads (e.g., required for biting) are transferred to the bone-implant interface enables better design of implant supported prosthesis as well [10], [11].

Traditional strain measurement technique in experiments uses strain gauges. Although they are known to be reliable and robust [12], strain gauges only give discrete data by averaging the real strains between the tips of strain gauge. The result of this technique tends to be affected by environment and requires preparation of the surface and wiring [13]. In general, local strain measurements are essential because complexity in geometries and inhomogeneity result in substantial variation over the structure. Recently, a technique named Digital Image Correlation (DIC) has been developed that captures full field strain data with a good resolution on the surface of the model from images. This relatively new method can be applied to complex geometries. Theories and detailed information of the method can be found in Refs. [14], [15]. This DIC technique is used extensively in experiments to characterize the specimen’s behavior. Using this method, full field strain of splinted and non-splinted implant prostheses has been captured [16]. Also, various kinds of computational models for composite hemipelvis [12] and, proximal femur [12], [17] have been validated with DIC.

The principal aim of this paper is twofold. (i) To experimentally validate that the finite element model correctly captures the stress strain contour of the skull model subjected to masticatory forces and (ii) to validate that the designed implant appropriately transfer the masticatory forces by keeping the maximum stress below ultimate stress. In order to do this, a bone replacement shape for a patient with massive facial injury with large bone loss in midface is designed using the topology optimization method. The solution is then embedded in the skull with computer-aided engineering software to obtain a skull model. Then, the skull model with the embedded implant is fabricated by a 3D printer. Mastication activity is simulated in the experiment and also modeled by finite element analysis (FEA). This work is expected not only to evaluate the solution of topology optimization but also to appraise the idea of using topology optimization to design macroscopic bone replacement shapes for midface defects. The full field strain data from mechanical testing via DIC is compared with that from FE (Finite element) model under same boundary and loading conditions to validate the FE model. FE model of skull is further analyzed to investigate the practicality of the skull in terms of load transfer mechanism and structural integrity.

Section snippets

Model preparation

A skull model of massive facial injury with large bone loss in the midface is extracted from the CT scan using software Amira (FEI Visualization Science Group, MA, USA) as shown in Fig. 1. The defect is in the center of midface and is asymmetrically extended bilaterally. Using measurement tools, the appropriate design domain is extracted. Within the design domain, supports are provided on the lateral surfaces for contact and fixation between the implant and uninjured portion of facial skeleton.

Results and discussion

The full field strains captured from mechanical testing on three skull models are compared with one another to ensure the reliability of the 3D printed skull models as well as the setup of the mechanical testing. Maximum and principal strains and vertical strain components in the area of interest (as shown in Fig. 4(a)) are calculated. The applied loading resulted in positive maximum principal strain (tensile), whereas minimum principal strain was negative (compressive) in most areas of the

Conclusion and future work

In this work, the mechanical feasibility of implants designed by the topology optimization method for craniofacial reconstructive surgery is examined. A mastication simulation using a finite element model that was generated from patient’s CT scan is validated with mechanical testing. The specimen for the mechanical testing was obtained using a 3D printer. The FEA results show that topology optimized solutions have the potential to restore destroyed buttress systems and recover the structural

Funding

National Science Foundation through the CMMI Award no. 1032884.

Conflict of interest statement

None.

Acknowledgement

The work was funded by National Science Foundation through the CMMI Award no. 1032884. Diana Carrau was supported by the MDRS Roessler fellowship from College of Medicine, The Ohio State University. We would like to acknowledge Dr. Jeremy Siedt, Department of Mechanical and Aerospace Engineering, The Ohio State University for facilitating the mechanical testing of the skull model and Professor Blaine Lilly and Mr. Richard Teynor for helping with the 3D printing. Finally, we thank Dr. Tam H

References (27)

  • A. Sutradhar et al.

    Computational modeling of craniofacial bone replacements

    Med. Biol. Eng. Comput.

    (2013)
  • A. Sutradhar et al.

    Topological optimization for designing patient-specific large craniofacial segmental bone replacements

    Proc. Nat. Acad. Sci. U.S.A.

    (2010)
  • L. Krog et al.

    Application of topology, sizing and shape optimization methods to optimal design of aircraft components, in: Product Design

    (2011)
  • Cited by (0)

    View full text