1 Introduction

Additive manufacturing (widely known as 3D printing) and rapid prototyping have found many applications in medicine [1, 2]. They are in use everywhere there is a need of an individualized solution, as they allow fast obtainment of a demanded shape with no tooling [3]. One of the main barriers of widespread production of professional 3D printing is the price of both the machines and materials [4, 5], as well as the required qualifications and time consumption to both prepare data and launch and supervise the additive manufacturing processes [4]. In recent years, 3D printing has gained wide popularity thanks to so-called low-cost additive manufacturing processes, which are mostly inexpensive variations of the long-known fused deposition modeling technology. However, they have certain limitations. Another problem is lack of standard methodologies of designing medical products on the basis of medical imaging data.

The mid-surgery supplies are an important aid for a surgeon during the stage of planning and scheduling a detailed course of operation, as well as during the operation itself, for example as a template for bone cutting [6, 7]. The main area in which the authors focus their work is surgery, especially in terms of reconstruction and implanting after resection. The use of 3D printed metal implants is nowadays a known technique [1], although limited in use. Use of low-cost 3D printing to produce mid-surgery supplies is uncommon, mostly due to lack of low-cost metal-processing techniques and process stability and product quality problems.

All of the above-mentioned problems are premises of research undertaken by the authors of this paper. The general and long-term aim of the research is to establish a reliable set of methods of building medical products of certain categories in order to make it more available for both doctors and patients.

2 Introduction

2.1 Research Problem

The main problem of this paper is a proper utilization of low-cost additive manufacturing and Rapid Prototyping techniques as tools of effective manufacturing of usable, individualized medical products, for use both by doctors and patients on various stages of treatment. This paper addresses issues of design and low-cost 3D printing on the basis of CT (computed tomography) data. The considerations are based on examples of specific cases of medical products used by doctors in real-world scenarios. As the main case, a mandibular reconstruction template was selected.

2.2 Methodology of Design and Rapid Manufacturing of Medical Products

The proposed methodology is based on other available general methodologies [2, 8]. However, the authors have expanded the existing methodologies with the aspects of clear indication of required skills and involved people and indication of proposed software at each stage, mostly aiming at freeware or open-source software.

The best and most suitable tools, as tested by the authors, are indicated in the full methodology, which is not included here due to limited space. The methodology scheme is presented in Fig. 1.

Fig. 1
figure 1

Schematic course of proposed methodology to design and manufacture medical products using 3D printing techniques

Full size image

The authors have distinguished eight consecutive stages that require at least three separate competence areas: medical technician (stages 1 and 8), biomedical engineer (stages 2–4) and 3D printing technician (stages 5–7). The most demanding stage in terms of skills and time is Stage 4, where a surface and/or solid model must be created. Stage 4 can be especially demanding when Stage 2 is conducted improperly [9]. Stage 3 also requires a large amount of skills of manual processing of 3D mesh.

2.3 The Fused Deposition Modelling Technology

The fused deposition modeling (FDM) process is currently the most widespread additive manufacturing technology [10]. The consistent problem of the FDM process is that produced parts have weak interlayer bonds, which decrease overall strength of the part [11]. Furthermore, part orientation greatly influences the obtained technical characteristics [12]. The problems of selecting optimal parameters of the FDM process and obtaining good properties of products has been widely described in literature [13].

For the studies, two machines were used—a Stratasys Dimension BST 1200 (a professional machine) and a MakerBot Replicator 2X (a low-cost machine). To provide a valid comparison for later evaluation, in the conducted studies similar parameters and materials were selected for both machines.

2.4 Mandibular Reconstruction Template—Case Description

The main case is mandibular reconstruction using fibular free flaps in patients after resections. Proper planning of the reconstruction surgery using 3D data of the lower jaw, obtained by medical imaging, has been widely described in literature [14]. Application of rapid prototyping methods in such a procedure is evaluated as effective and helpful, especially in terms of reduction of operation time and better geometrical fitting of the reconstructed jaw [15]. However, the costs of such an approach are high.

The case presented in this paper involves a patient subjected to partial lower jaw resection. The patient was male and aged 47. The side part of the jaw body had to be removed due to cancer and the resulting defect was reconstructed. It was aimed at obtaining full jaw functionality, including junction stability, mobility and cosmetic effect. The FDM process was used to manufacture a template of the reconstructed jaw out of ABS material, to use before and in the middle of the surgery, to shape stabilizing titanium plates (Fig. 2) and to be used as a template to cut the desired shape from patient’s fibula in shorter time. The approach was tested earlier by the authors in clinical tests, using solely the professional BST 1200 machine, with full success.

Fig. 2
figure 2

Example of fitting of titanium plates on a lower jaw template, before the surgery

Full size image

2.5 Course of Studies

The studies were divided into two main parts. The first part focused on using the defined methodology to manufacture the jaw templates using both low-cost and professional FDM processes out of ABS material, starting from medical data processing, through design, 3D printing process preparation and post-processing. The second part of the studies focused on evaluation of the obtained products.

Firstly, medical imaging data was obtained, in form of DICOM file with CT data (Stage 1). The computed tomography images of lower jaws were converted to digital 3D models (Stage 2). The created models were exported to the STL format. In further stages, the raw models were processed in the GOM Inspect software, until final shape was obtained (Stage 3). In a similar way, a digital model of the fibula was obtained. Models prepared in such a way were imported to a computer-aided design system. A piece of damaged bone was removed from the lower jaw model. Then, a missing piece was digitally recreated (Stage 4). In the next stage (Stage 5), the FDM process was prepared. Similar manufacturing parameters and materials were used (see Table 1). In stages 6 and 7, the jaw templates were manufactured and post processed. The post-processing was the same in both cases and consisted solely in mechanical support removal.

Table 1 Comparison of basic information about manufacturing on two FDM machines
Full size table

At this point, both products were optically measured using Atos Compact Scan 5M 3D scanner. Measurement field was set to 130 × 150 mm. Both scans were superimposed on the original, final STL export (nominal geometry), using the GOM Inspect software and the best-fit algorithm. Inspection points locations were consulted with the surgeons.

3 Results

The manufactured reconstructed mandible templates are shown in Fig. 3. Basic information regarding the manufacturing processes and obtained economical characteristics are presented in Table 1.

Fig. 3
figure 3

3D printed jaw templates, Replicator 2X (left), Dimension BST 1200 (right)

Full size image

The 3D scanning and inspection results are presented in Fig. 4. The average scan fitting error was 0.05 mm for the BST 1200 machine and 0.27 mm for the Replicator 2X machine. In terms of fitting of the selected 11 points—for the BST 1200 machine all points are within tolerance of ±0.05 mm, while for the Replicator 2X all points are outside this tolerance, with maximal deviation −0.56 mm in point no. 5.

Fig. 4
figure 4

Results of accuracy inspection—colorful deviation map, Replicator 2X model (left) and Dimension BST 1200 model (right)

Full size image

Both products were sterilized in plasma at 55 °C for 45 min. No geometrical changes were observed. Both products were then evaluated and compared by a team of surgeons. No visible quality defects were observed and the surgeons marked both products as acceptable for use during the surgery. The surgery was performed and it ended with full success and patient recovery.

4 Discussion

In terms of product design issues, it was found that it is required to use at most three different software tools in order to generate the mesh out of medical data, process the mesh and create a design. The most labor-consuming process is the work done by a biomedical engineer, i.e. time spent in-between raw DICOM data and a final computer-aided model. In the presented case, the time of work was 12 h. This is, in authors’ opinion, a rather short time, with a standard time of 3–6 days. The paper describes a single case, but the methodology was applied to several other cases as well, with similar results. Still, more clinical tests would be required, but this type of surgery is rather rare, so it will probably take a few more years.

Out of the medical product classes available nowadays on low-cost 3D printers, the mid-surgery supplies production using low-cost 3D printing is the most viable option in terms of usefulness and cost efficiency, in authors’ opinion. Its effectiveness has been confirmed by clinical tests.

A low-cost FDM process allows the possibility to reduce product costs by more than two times, but at the cost of greatly limited dimensional accuracy. Fortunately, in terms of products that resemble the shape of the human body, the accuracy ensured by low-cost FDM is enough, as proven by the positive results of the performed surgery (contrary to findings by other researchers, such as in [9]. Unfortunately, low-cost FDM processes require a considerably high level of supervision. Examples of errors are clogging of extruders, miscalibrated building plate, filament entanglement and self-unsticking of the model. Each error requires operator to perform manual actions. It is, in authors’ opinion, a large barrier in making the 3D printing widespread in healthcare.

5 Conclusions

Based upon the presented studies, the following final conclusions may be drawn:

  1. 1.

    It is effectively possible to use a low-cost 3D printer to manufacture a medical product usable by surgeons during an operation, on the basis of computed tomography data.

  2. 2.

    Use of a defined methodology shortens time spent on searching for appropriate tools and ways of conduct, thus reducing time of preparation of a medical product, which is often a crucial factor in life-saving surgery.

  3. 3.

    Obtaining usable medical products by methods of 3D printing requires cooperation between at least 4 specialists: a doctor, a medical imaging technician, a biomedical engineer (computer-aided design specialist) and a 3D printing technician.

Future work will focus on more clinical tests of different medical products, such as prostheses, orthoses and mid-surgery aids, as well as developing communication standards between engineers and doctors.