Design and manufacture of customized dental implants by using reverse engineering and selective laser melting technology Jianyu Chen, PhD,a Zhiguang Zhang, MD,b Xianshuai Chen, PhD,c Chunyu Zhang, MD,d Gong Zhang, PhD,e and Zhewu Xu, PhDf Sun Yat-sen University, Guangzhou, China; Chinese Academy of Science, Guangzhou, China Statement of problem. Recently a new therapeutic concept of patient-specific implant dentistry has been advanced based on computer-aided design/computer-aided manufacturing technology. However, a comprehensive study of the design and 3-dimensional (3D) printing of the customized implants, their mechanical properties, and their biomechanical behavior is lacking. Purpose. The purpose of this study was to evaluate the mechanical and biomechanical performance of a novel custom-made dental implant fabricated by the selective laser melting technique with simulation and in vitro experimental studies. Material and methods. Two types of customized implants were designed by using reverse engineering: a root-analog implant and a root-analog threaded implant. The titanium implants were printed layer by layer with the selective laser melting technique. The relative density, surface roughness, tensile properties, bend strength, and dimensional accuracy of the specimens were evaluated. Nonlinear and linear finite element analysis and experimental studies were used to investigate the stress distribution, micromotion, and primary stability of the implants. Results. Selective laser melting 3D printing technology was able to reproduce the customized implant designs and produce high density and strength and adequate dimensional accuracy. Better stress distribution and lower maximum micromotions were observed for the root-analog threaded implant model than for the root-analog implant model. In the experimental tests, the implant stability quotient and pull-out strength of the 2 types of implants indicated that better primary stability can be obtained with a root-analog threaded implant design. Conclusions. Selective laser melting proved to be an efficient means of printing fully dense customized implants with high strength and sufficient dimensional accuracy. Adding the threaded characteristic to the customized root-analog threaded implant design maintained the approximate geometry of the natural root and exhibited better stress distribution and primary stability. (J Prosthet Dent 2014;112:1088-1095)
Clinical Implications Investigating the stress distribution patterns and the primary stability of 2 implant designs with different loading conditions provides new information about the design and 3-dimensional (3D) printing of customized dental implants.
Supported by grant No. 81271115 from the Natural Science Foundation of China. a
Resident, Guanghua School of Stomatology, Hospital of Stomatology, Guangdong Provincial Key Laboratory of Stomatology, Sun Yat-sen University. b Professor, Department of Oral and Maxillofacial Surgery, Guanghua School of Stomatology, Hospital of Stomatology, Sun Yat-sen University. c Associate Professor, Guangzhou Institute of Advanced Technology, Chinese Academy of Science. d Engineer, Guangzhou Institute of Advanced Technology, Chinese Academy of Science. e Engineer, Guangzhou Institute of Advanced Technology, Chinese Academy of Science. f Resident, Guanghua School of Stomatology, Hospital of Stomatology, Sun Yat-sen University.
The Journal of Prosthetic Dentistry
Chen et al
November 2014 Since the concept of osseointegration was introduced by Brånemark et al,1 modern dental implants have been considered a safe and reliable option for replacing missing teeth.2 Because of the individual alveolar bone phenotype characteristics of each patient3 and the wish for more precise surgery and earlier function, a new therapeutic concept based on patient-specific implant dentistry has been suggested.4-12 Current commercial dental implants have cylindrical or tapered geometries with threads along their length.13 However, commercially standard dental implants provide only limited options for implant length, diameter, and thread parameters and cannot completely meet the requirements for all individual oral conditions.3,5,11 Thus, customized dental implants tailored to individual patients not only preserve more hard and soft tissues but also reduce rehabilitation time, opening a promising prospective for implant dentistry.7,9 With the rapid development of implant materials and computer-aided design/ computer-aided manufacturing (CAD/ CAM) technology, many researchers have recently reconsidered the customized implant to tackle the incongruity between conventional implant geometry and available bone.4,8,9,14,15 The successful clinical uses of modified rootanalog zirconia implants fabricated by CAD/CAM technology for immediate tooth replacement have been reported, and excellent esthetic and functional results have been achieved with minimal bone resorption and soft tissue recession at a 2.5 year follow-up.9 Dental implants closely imitating natural root geometries are difficult to fabricate with conventional methods of computer numerical control machining because of the existence of irregularly curved surfaces and complex 3-dimensional (3D) geometry.4 Additive manufacturing (AM), also known as 3D printing, is capable of directly producing almost any desired geometry without an expensive mold and tooling. AM has been widely considered as the future of custom-made implants and of
Chen et al
1089 research.16 AM methods for metals can be divided into 2 main groups: selective laser sintering (SLS) or direct metal laser sintering (DMLS), where the surface of metal particles is partially melted to join them in a more or less porous structure, and selective laser melting (SLM), where the metal particles are completely melted and form a dense metal from locally produced volumes of liquid metal. So far, only SLS/DMLS technology has been reported for the fabrication of dental implants.6,17 However, these dental implants exhibited apparent defects: rough surface, pores, cracks, and distortion in threads were caused by the intrinsic nature of the laser sintering process in SLS, including the partial melting and superficial fusing of metal powder. From an engineering perspective, pores and defects in the material as starting points for cracks may also have a negative effect on the mechanical properties of the dental implants under long-term cyclic loading.15 SLM is an advanced AM technique that can fabricate dense parts by completely melting metal powder to produce material properties close to those of the bulk materials.18,19 No comprehensive studies of geometric design and SLMprocessed customized dental implants have yet been reported. Hence, the purpose of this study was to describe a novel custom-made dental implant with a threaded characteristic produced by the SLM method. The mechanical properties and biomechanical performance of the customized implants printed by SLM technology were evaluated by both simulation analysis and in vitro experimental studies. The null hypothesis was that implants with the threaded characteristic would not differ significantly from those without it.
root-analog implant (RI) and a rootanalog threaded implant (RTI), were created from this incisor by using CAD software (SolidWorks v12.0; Dassault Systèmes SolidWorks Corp). The model of the RI was created according to the anatomic root morphology. The model of the RTI was created by fitting a rotationally symmetric body to the natural root morphology. Because of the curvature and asymmetric geometry of the incisor, the least square method was used to seek a compromise between the natural geometry of the dental alveolus and the geometry of the rotationally symmetric RTI (see Supplemental Material). A CAD model of a simplified anterior maxillary bone block with a 1-mm layer of cortical bone was created based on the CT images. The edentulous bone block with a simulated extraction dental alveolus was obtained by removing the maxillary incisor. The customized implants were spatially embedded into the alveolus, and a minimal 1-mm labial plate was guaranteed according to the principle of the immediate implant surgery. For the simulations of immediate loading after implant placement, nonlinear surfaceto-surface contact elements were adopted in the implant-bone interface (m¼0.45).20,21 To simulate the loading of osseointegrated implants, a bonded condition was set at the implant-bone interface.21 All materials used in the simulation were considered isotropic, homogeneous, and linearly elastic.22,23 The mechanical properties10 used for the simulation are shown in Table I.
Table I. Material properties used in finite element models
MATERIAL AND METHODS Materials A series of computed tomography (CT) images of a maxillary incisor was selected and input into software (Mimics v10.01; Materialise) to obtain a model of an incisor. The CAD models of the customized dental implants, a
Cortical bone
Young Modulus Poisson E (MPa) Ratio n 13 800
0.26
345
0.31
Titanium
110 000
0.35
Porcelain
70 000
0.19
Cancellous bone
1090
Volume 112 Issue 5 used to measure the implant stability quotient (ISQ) of each implant, and each measurement was repeated 3 times. After the measurement of the resonance frequency, all implants were extracted from the sockets at a rate of 30 mm/min in line with the implant axis by using a universal testing machine (KD-III; Kaiqiangli Testing Instruments Co Ltd) until the implants were completely removed from the artificial bone. The maximum force during this process was recorded as the pull-out strength of each implant.
RESULTS 1 Computer-aided design model of root-analog implant (direction of occlusal force indicated by red arrow).
2 Computer-aided design model of root-analog threaded implant (direction of occlusal force indicated by red arrow).
The surface of the bone segment opposite the alveolus was treated as a fixed boundary. To simulate the natural occlusal force, a static load of 100 N was applied to the simplified porcelain superstructure (Figs. 1, 2).24 Commercial grade 2 pure Ti powder was used as a starting material for the dental implants. An SLM machine featuring a diode-pumped ytterbium fiber laser and laser optics that produce a focal point of about 90 mm in diameter (SLM125HL; SLM solutions GmbH) was used to fabricate the dental implants. The SLM laser processing parameters were optimized as a laser power of 100 W, laser scanning velocity of 275 mm/s, hatch distance of 130 mm, and layer thickness of 30 mm. To prevent oxidation of the Ti powder, the SLM process was run under argon with a purity of 99.999 wt%. The images of the custom-made implants were made with a stereomicroscope (Leica M205A; Leica Microsystems GmbH) after airborne-particle abrasion. The densities of the dental implants were measured with the Archimedes principle, and the relative densities were determined by ractual/rtheoretical. The surface roughness and the dimensional accuracy of the major implant features, including diameter and length, were measured. To determine the strength of
SLM-processed Ti material, standard specimens for tensile testing and bending testing were directly produced by the SLM machine. Tensile tests were carried out with a mechanical testing machine (SANS-CMT5105; MTS System Co) according to the International Organization for Standardization (ISO) standard ISO 6892. Three-point bending tests were performed according to ISO 7438 by using a mechanical testing machine (Instron-E300; Instron Co). The implant specimens fabricated by the SLM machine (10 each of RI and RTI) were used in the primary stability test. Mechanical test blocks of artificial bone (Sawbones; Pacific Research Laboratories Inc) with 2 corresponding bone mineral densities (0.16 g/cm3 for group A and 0.24 g/cm3 for group B) were selected as a jaw bone equivalent. Cavities in the bone blocks were prepared by a milling machine to simulate the postextraction alveolus. The RIs were placed into the alveolus by 1 technician.7,12 The RTIs were screwed into the cavities manually with a hand driver after the pilot holes had been prepared step by step with a dental drill (first drill: ⌀¼1.0 mm, length¼14 mm; second drill: ⌀¼2.5 mm, length¼9 mm; third drill: ⌀¼3.7 mm, length¼5 mm). A resonance frequency analyzer (Osstell ISQ; Integration Diagnostics Ltd) was
The Journal of Prosthetic Dentistry
When the implants were loaded with oblique occlusal force immediately, the maximum von Mises stress (EQV) located in the cortical bone-implant interface of the RTI was about 42%, lower than that of the RI. The stress for the RTI was evenly distributed around the implant neck, and the peak stress was concentrated in the palatal cortical bone with a value of 66.7 MPa. The peak stress for the RI was observed in the irregular curved contour. The RI model showed a maximum EQV distribution of the cancellous bone near the implant apex (Fig. 3). The peak stress observed in the interface between the cancellous bone and the RTI was located in the palatal first thread (Fig. 4). The maximum micromotions between the implant-bone interfaces were found in the cortical bone, with a value of 102.7 mm for the RI and 56.4 mm for the RTI (Table II). With regard to the osseointegrated implants, the general patterns for stress distribution were relatively similar for the 2 designs. The tension stresses in the palatal sides and the compression stresses in the labial sides were observed in the crestal regions, with a maximum value of 53.8 MPa for the RI and 52.9 MPa for the RTI. The stress magnitude in the cancellous bone appeared to be minor because of its lower elastic modulus. A stress concentration was observed around the apical and middle section of the implant in labial cancellous bone, with a maximum value of 3.9 MPa for the
Chen et al
November 2014
1091
3 Stress distributions of immediate loading protocol: root-analog implant. RI and 3.4 MPa for the RTI (Figs. 5, 6). The maximum micromotion at the implant-bone interfaces for the bonded contact models was also similar, around 32 mm for the RI and RTI (see Table II). The images of the customized dental implants are shown in Figures 7 and 8. Although the surfaces display a macroroughness unique to the SLM process, they were free of distortion, pores, and cracks. The relative density, surface roughness, yield strength, bending strength, and dimensional accuracy of the dental implants are presented in Table III. Figure 9 shows the mean ISQ value and the standard deviations of the 2 customized implants. One-way ANOVA confirmed that the ISQ values of the RTI inserted into the artificial bones were significantly higher than those of
Chen et al
4 Stress distributions of immediate loading protocol: root-analog threaded implant.
the RI (P
Clinical Implications Investigating the stress distribution patterns and the primary stability of 2 implant designs with different loading conditions provides new information about the design and 3-dimensional (3D) printing of customized dental implants.
Supported by grant No. 81271115 from the Natural Science Foundation of China. a
Resident, Guanghua School of Stomatology, Hospital of Stomatology, Guangdong Provincial Key Laboratory of Stomatology, Sun Yat-sen University. b Professor, Department of Oral and Maxillofacial Surgery, Guanghua School of Stomatology, Hospital of Stomatology, Sun Yat-sen University. c Associate Professor, Guangzhou Institute of Advanced Technology, Chinese Academy of Science. d Engineer, Guangzhou Institute of Advanced Technology, Chinese Academy of Science. e Engineer, Guangzhou Institute of Advanced Technology, Chinese Academy of Science. f Resident, Guanghua School of Stomatology, Hospital of Stomatology, Sun Yat-sen University.
The Journal of Prosthetic Dentistry
Chen et al
November 2014 Since the concept of osseointegration was introduced by Brånemark et al,1 modern dental implants have been considered a safe and reliable option for replacing missing teeth.2 Because of the individual alveolar bone phenotype characteristics of each patient3 and the wish for more precise surgery and earlier function, a new therapeutic concept based on patient-specific implant dentistry has been suggested.4-12 Current commercial dental implants have cylindrical or tapered geometries with threads along their length.13 However, commercially standard dental implants provide only limited options for implant length, diameter, and thread parameters and cannot completely meet the requirements for all individual oral conditions.3,5,11 Thus, customized dental implants tailored to individual patients not only preserve more hard and soft tissues but also reduce rehabilitation time, opening a promising prospective for implant dentistry.7,9 With the rapid development of implant materials and computer-aided design/ computer-aided manufacturing (CAD/ CAM) technology, many researchers have recently reconsidered the customized implant to tackle the incongruity between conventional implant geometry and available bone.4,8,9,14,15 The successful clinical uses of modified rootanalog zirconia implants fabricated by CAD/CAM technology for immediate tooth replacement have been reported, and excellent esthetic and functional results have been achieved with minimal bone resorption and soft tissue recession at a 2.5 year follow-up.9 Dental implants closely imitating natural root geometries are difficult to fabricate with conventional methods of computer numerical control machining because of the existence of irregularly curved surfaces and complex 3-dimensional (3D) geometry.4 Additive manufacturing (AM), also known as 3D printing, is capable of directly producing almost any desired geometry without an expensive mold and tooling. AM has been widely considered as the future of custom-made implants and of
Chen et al
1089 research.16 AM methods for metals can be divided into 2 main groups: selective laser sintering (SLS) or direct metal laser sintering (DMLS), where the surface of metal particles is partially melted to join them in a more or less porous structure, and selective laser melting (SLM), where the metal particles are completely melted and form a dense metal from locally produced volumes of liquid metal. So far, only SLS/DMLS technology has been reported for the fabrication of dental implants.6,17 However, these dental implants exhibited apparent defects: rough surface, pores, cracks, and distortion in threads were caused by the intrinsic nature of the laser sintering process in SLS, including the partial melting and superficial fusing of metal powder. From an engineering perspective, pores and defects in the material as starting points for cracks may also have a negative effect on the mechanical properties of the dental implants under long-term cyclic loading.15 SLM is an advanced AM technique that can fabricate dense parts by completely melting metal powder to produce material properties close to those of the bulk materials.18,19 No comprehensive studies of geometric design and SLMprocessed customized dental implants have yet been reported. Hence, the purpose of this study was to describe a novel custom-made dental implant with a threaded characteristic produced by the SLM method. The mechanical properties and biomechanical performance of the customized implants printed by SLM technology were evaluated by both simulation analysis and in vitro experimental studies. The null hypothesis was that implants with the threaded characteristic would not differ significantly from those without it.
root-analog implant (RI) and a rootanalog threaded implant (RTI), were created from this incisor by using CAD software (SolidWorks v12.0; Dassault Systèmes SolidWorks Corp). The model of the RI was created according to the anatomic root morphology. The model of the RTI was created by fitting a rotationally symmetric body to the natural root morphology. Because of the curvature and asymmetric geometry of the incisor, the least square method was used to seek a compromise between the natural geometry of the dental alveolus and the geometry of the rotationally symmetric RTI (see Supplemental Material). A CAD model of a simplified anterior maxillary bone block with a 1-mm layer of cortical bone was created based on the CT images. The edentulous bone block with a simulated extraction dental alveolus was obtained by removing the maxillary incisor. The customized implants were spatially embedded into the alveolus, and a minimal 1-mm labial plate was guaranteed according to the principle of the immediate implant surgery. For the simulations of immediate loading after implant placement, nonlinear surfaceto-surface contact elements were adopted in the implant-bone interface (m¼0.45).20,21 To simulate the loading of osseointegrated implants, a bonded condition was set at the implant-bone interface.21 All materials used in the simulation were considered isotropic, homogeneous, and linearly elastic.22,23 The mechanical properties10 used for the simulation are shown in Table I.
Table I. Material properties used in finite element models
MATERIAL AND METHODS Materials A series of computed tomography (CT) images of a maxillary incisor was selected and input into software (Mimics v10.01; Materialise) to obtain a model of an incisor. The CAD models of the customized dental implants, a
Cortical bone
Young Modulus Poisson E (MPa) Ratio n 13 800
0.26
345
0.31
Titanium
110 000
0.35
Porcelain
70 000
0.19
Cancellous bone
1090
Volume 112 Issue 5 used to measure the implant stability quotient (ISQ) of each implant, and each measurement was repeated 3 times. After the measurement of the resonance frequency, all implants were extracted from the sockets at a rate of 30 mm/min in line with the implant axis by using a universal testing machine (KD-III; Kaiqiangli Testing Instruments Co Ltd) until the implants were completely removed from the artificial bone. The maximum force during this process was recorded as the pull-out strength of each implant.
RESULTS 1 Computer-aided design model of root-analog implant (direction of occlusal force indicated by red arrow).
2 Computer-aided design model of root-analog threaded implant (direction of occlusal force indicated by red arrow).
The surface of the bone segment opposite the alveolus was treated as a fixed boundary. To simulate the natural occlusal force, a static load of 100 N was applied to the simplified porcelain superstructure (Figs. 1, 2).24 Commercial grade 2 pure Ti powder was used as a starting material for the dental implants. An SLM machine featuring a diode-pumped ytterbium fiber laser and laser optics that produce a focal point of about 90 mm in diameter (SLM125HL; SLM solutions GmbH) was used to fabricate the dental implants. The SLM laser processing parameters were optimized as a laser power of 100 W, laser scanning velocity of 275 mm/s, hatch distance of 130 mm, and layer thickness of 30 mm. To prevent oxidation of the Ti powder, the SLM process was run under argon with a purity of 99.999 wt%. The images of the custom-made implants were made with a stereomicroscope (Leica M205A; Leica Microsystems GmbH) after airborne-particle abrasion. The densities of the dental implants were measured with the Archimedes principle, and the relative densities were determined by ractual/rtheoretical. The surface roughness and the dimensional accuracy of the major implant features, including diameter and length, were measured. To determine the strength of
SLM-processed Ti material, standard specimens for tensile testing and bending testing were directly produced by the SLM machine. Tensile tests were carried out with a mechanical testing machine (SANS-CMT5105; MTS System Co) according to the International Organization for Standardization (ISO) standard ISO 6892. Three-point bending tests were performed according to ISO 7438 by using a mechanical testing machine (Instron-E300; Instron Co). The implant specimens fabricated by the SLM machine (10 each of RI and RTI) were used in the primary stability test. Mechanical test blocks of artificial bone (Sawbones; Pacific Research Laboratories Inc) with 2 corresponding bone mineral densities (0.16 g/cm3 for group A and 0.24 g/cm3 for group B) were selected as a jaw bone equivalent. Cavities in the bone blocks were prepared by a milling machine to simulate the postextraction alveolus. The RIs were placed into the alveolus by 1 technician.7,12 The RTIs were screwed into the cavities manually with a hand driver after the pilot holes had been prepared step by step with a dental drill (first drill: ⌀¼1.0 mm, length¼14 mm; second drill: ⌀¼2.5 mm, length¼9 mm; third drill: ⌀¼3.7 mm, length¼5 mm). A resonance frequency analyzer (Osstell ISQ; Integration Diagnostics Ltd) was
The Journal of Prosthetic Dentistry
When the implants were loaded with oblique occlusal force immediately, the maximum von Mises stress (EQV) located in the cortical bone-implant interface of the RTI was about 42%, lower than that of the RI. The stress for the RTI was evenly distributed around the implant neck, and the peak stress was concentrated in the palatal cortical bone with a value of 66.7 MPa. The peak stress for the RI was observed in the irregular curved contour. The RI model showed a maximum EQV distribution of the cancellous bone near the implant apex (Fig. 3). The peak stress observed in the interface between the cancellous bone and the RTI was located in the palatal first thread (Fig. 4). The maximum micromotions between the implant-bone interfaces were found in the cortical bone, with a value of 102.7 mm for the RI and 56.4 mm for the RTI (Table II). With regard to the osseointegrated implants, the general patterns for stress distribution were relatively similar for the 2 designs. The tension stresses in the palatal sides and the compression stresses in the labial sides were observed in the crestal regions, with a maximum value of 53.8 MPa for the RI and 52.9 MPa for the RTI. The stress magnitude in the cancellous bone appeared to be minor because of its lower elastic modulus. A stress concentration was observed around the apical and middle section of the implant in labial cancellous bone, with a maximum value of 3.9 MPa for the
Chen et al
November 2014
1091
3 Stress distributions of immediate loading protocol: root-analog implant. RI and 3.4 MPa for the RTI (Figs. 5, 6). The maximum micromotion at the implant-bone interfaces for the bonded contact models was also similar, around 32 mm for the RI and RTI (see Table II). The images of the customized dental implants are shown in Figures 7 and 8. Although the surfaces display a macroroughness unique to the SLM process, they were free of distortion, pores, and cracks. The relative density, surface roughness, yield strength, bending strength, and dimensional accuracy of the dental implants are presented in Table III. Figure 9 shows the mean ISQ value and the standard deviations of the 2 customized implants. One-way ANOVA confirmed that the ISQ values of the RTI inserted into the artificial bones were significantly higher than those of
Chen et al
4 Stress distributions of immediate loading protocol: root-analog threaded implant.
the RI (P
