Christoph Kunz

Medical image segmentation, whether semi-automatically or manually, is labor-intensive, subjective, and needs specialized personnel. The fully automated segmentation process recently gained importance due to its better design and understanding of CNNs. Considering this, we decided to develop our in-house segmentation software and compare it to the systems of established companies, an inexperienced user, and an expert as ground truth. The companies included in the study have a cloud-based option that performs accurately in clinical routine (dice similarity coefficient of 0.912 to 0.949) with an average segmentation time ranging from 3′54″ to 85′54″. Our in-house model achieved an accuracy of 94.24% compared to the best-performing software and had the shortest mean segmentation time of 2′03″. During the study, developing in-house segmentation software gave us a glimpse into the strenuous work that companies face when offering clinically relevant solutions. All the problems encountered were discussed with the companies and solved, so both parties benefited from this experience. In doing so, we demonstrated that fully automated segmentation needs further research and collaboration between academics and the private sector to achieve full acceptance in clinical routines.

Distraction osteogenesis is a clinically established technique for lengthening, molding and shaping bone by new bone formation. The experimental evaluation of this expensive and time-consuming treatment is of high impact for better understanding of tissue engineering but mainly relies on a limited number of histological slices. These tissue slices contain two-dimensional information comprising only about one percent of the volume of interest. In order to analyze the soft and hard tissues of the entire jaw of a single rat in a multimodal assessment, we combined micro computed tomography (µCT) with histology. The µCT data acquired before and after decalcification were registered to determine the impact of decalcification on local tissue shrinkage. Identification of the location of the H&E-stained specimen within the synchrotron radiation-based µCT data collected after decalcification was achieved via non-rigid slice-to-volume registration. The resulting bi- and tri-variate histograms were divided into clusters related to anatomical features from bone and soft tissues, which allowed for a comparison of the approaches and resulted in the hypothesis that the combination of laboratory-based µCT before decalcification, synchrotron radiation-based µCT after decalcification and histology with hematoxylin-and-eosin staining could be used to discriminate between different types of collagen, key components of new bone formation.

Individual cutting guides for the reconstruction of lower jaw defects with fibular grafts are often used. However, the application of these osteotomy tools is costly and time intensive. The aim of this study was to compare the precision of osteotomies using a 3D-printed guide with those using a universal, reusable, and more cost-efficient Multi-Use Cutting Jig (MUC-Jig). In this non-blinded experimental study, 10 cranio-maxillofacial surgeons performed four graft removals each in a randomized order using the same osteotomy angle, both proximally (sagittal cut) and distally (coronal cut), of a graft (45°, 30°, 15°, or 0°), first with the MUC-Jig then with the 3D-printed cutting guide. The 40 fibula transplants (Tx) of each method (n = 80) were then analyzed concerning their Tx length and osteotomy angles and compared to the original planning data. Furthermore, the surgeons’ subjective perception and the duration of the two procedures were analyzed. The mean relative length and mean relative angle deviation between the MUC-Jig (−0.08 ± 1.12 mm; −0.69° ± 3.15°) and the template (0.22 ± 0.90 mm; 0.36° ± 2.56°) group differed significantly (p = 0.002; p = < 0.001), but the absolute deviations did not (p = 0.206; p = 0.980). Consequently, clinically comparable osteotomy results can be achieved with both methods, but from an economic point of view the MUC-Jig is a more cost-efficient solution.

Additive manufacturing (AM) of patient-specific implants (PSIs) is gradually moving towards in-house or point-of-care (POC) manufacturing. Polyetheretherketone (PEEK) has been used in cranioplasty cases as a reliable alternative to other alloplastic materials. As only a few fused filament fabrication (FFF) printers are suitable for in-house manufacturing, the quality characteristics of the implants fabricated by FFF technology are still under investigated. This paper aimed to investigate PEEK PSIs fabricated in-house for craniofacial reconstruction, discussing the key challenges during the FFF printing process. Two exemplary cases of class III (Group 1) and class IV (Group 2) craniofacial defects were selected for the fabrication of PEEK PSIs. Taguchi’s L9 orthogonal array was selected for the following nonthermal printing process parameters, i.e., layer thickness, infill rate, number of shells, and infill pattern, and an assessment of the dimensional accuracy of the fabricated implants was made. The root mean square (RMS) values revealed higher deviations in Group 1 PSIs (0.790 mm) compared to Group 2 PSIs (0.241 mm). Horizontal lines, or the characteristic FFF stair-stepping effect, were more perceptible across the surface of Group 1 PSIs. Although Group 2 PSIs revealed no discoloration, Group 1 PSIs displayed different zones of crystallinity. These results suggest that the dimensional accuracy of PSIs were within the clinically acceptable range; however, attention must be paid towards a requirement of optimum thermal management during the printing process to fabricate implants of uniform crystallinity.

3D printed biomaterials have been extensively investigated and developed in the field of bone regeneration related to clinical issues. However, specific applications of 3D printed biomaterials in different dental areas have seldom been reported. In this study, we aimed to and successfully fabricated 3D poly (lactic-co-glycolic acid)/β-tricalcium phosphate (3D-PLGA/TCP) and 3D β-tricalcium phosphate (3D-TCP) scaffolds using two relatively distinct 3D printing (3DP) technologies. Conjunctively, we compared and investigated mechanical and biological responses on human dental pulp stem cells (hDPSCs). Physicochemical properties of the scaffolds, including pore structure, chemical elements, and compression modulus, were characterized. hDPSCs were cultured on scaffolds for subsequent investigations of biocompatibility and osteoconductivity. Our findings indicate that 3D printed PLGA/TCP and β-tricalcium phosphate (β-TCP) scaffolds possessed a highly interconnected and porous structure. 3D-TCP scaffolds exhibited better compressive strength than 3D-PLGA/TCP scaffolds, while the 3D-PLGA/TCP scaffolds revealed a flexible mechanical performance. The introduction of 3D structure and β-TCP components increased the adhesion and proliferation of hDPSCs and promoted osteogenic differentiation. In conclusion, 3D-PLGA/TCP and 3D-TCP scaffolds, with the incorporation of hDPSCs as a personalized restoration approach, has a prospective potential to repair minor and critical bone defects in oral and maxillofacial surgery, respectively.

The aim of this study was to compare the efficacy of the intraoperative bending of titanium mesh with the efficacy of pre-contoured “hybrid” patient-specific titanium mesh for the surgical repair of isolated orbital floor fractures. In-house 3D-printed anatomical models were used as bending guides. The main outcome measures were preoperative and postoperative orbital volume and surgery time. We performed a retrospective cohort study including 22 patients who had undergone surgery between May 2016 and November 2018. The first twelve patients underwent conventional reconstruction with intraoperative free-hand bending of an orbital floor mesh plate. The subsequent ten patients received pre-contoured plates based on 3D-printed orbital models that were produced by mirroring the non-fractured orbit of the patient using a medical imaging software. We compared the preoperative and postoperative absolute volume difference (unfractured orbit, fractured orbit), the fracture area, the fracture collapse, and the effective surgery time between the two groups. In comparison to the intraoperative bending of titanium mesh, the application of preformed plates based on a 3D-printed orbital model resulted in a non-significant absolute volume difference in the intervention group (p = 0.276) and statistically significant volume difference in the conventional group (p = 0.002). Further, there was a significant reduction of the surgery time (57.3 ± 23.4 min versus 99.8 ± 28.9 min, p = 0.001). The results of this study suggest that the use of 3D-printed orbital models leads to a more accurate reconstruction and a time reduction during surgery.

Computer-assisted surgery with three-dimensional (3D) printed surgical guides provides more accurate results than free-hand surgery. Steam sterilization could be one of the factors that affect the dimensions of surgical guide resin materials, leading to inaccuracies during surgeries. The purpose of this study was to evaluate the effects of steam sterilization on the dimensional accuracy of indication-specific hollow cube test bodies, manufactured in-house using Class IIa biocompatible resin materials (proprietary and third-party). To evaluate the pre- and post-sterilization dimensional accuracy, root mean square (RMS) values were calculated. The results indicate that, in all the groups, steam sterilization resulted in an overall linear expansion of the photopolymeric resin material, with an increase in outer dimensions and a decrease in inner dimensions. The effects on the dimensional accuracy of test bodies were not statistically significant in all the groups, except PolyJet Glossy (p > 0.05). The overall pre- and post-sterilization RMS values were below 100 and 200 µm, respectively. The highest accuracies were seen in proprietary resin materials, i.e., PolyJet Glossy and SLA-LT, in pre- and post-sterilization measurements, respectively. The dimensional accuracy of third-party resin materials, i.e., SLA-Luxa and SLA-NextDent, were within a comparable range as proprietary materials and can serve as an economical alternative.

The use of patient-specific implants (PSIs) in craniofacial surgery is often limited due to a lack of expertise and/or production costs. Therefore, a simple and cost-efficient template-based fabrication workflow has been developed to overcome these disadvantages. The aim of this study is to assess the accuracy of PSIs made from their original templates. For a representative cranial defect (CRD) and a temporo-orbital defect (TOD), ten PSIs were made from polymethylmethacrylate (PMMA) using computer-aided design (CAD) and three-dimensional (3D) printing technology. These customized implants were measured and compared with their original 3D printed templates. The implants for the CRD revealed a root mean square (RMS) value ranging from 1.128 to 0.469 mm with a median RMS (Q1 to Q3) of 0.574 (0.528 to 0.701) mm. Those for the TOD revealed an RMS value ranging from 1.079 to 0.630 mm with a median RMS (Q1 to Q3) of 0.843 (0.635 to 0.943) mm. This study demonstrates that a highly precise duplication of PSIs can be achieved using this template-molding workflow. Thus, virtually planned implants can be accurately transferred into haptic PSIs. This workflow appears to offer a sophisticated solution for craniofacial reconstruction and continues to prove itself in daily clinical practice.