Very few surgical teams currently use totally independent and free solutions to perform three-dimensional (3D) surgical modelling for osseous free flaps in reconstructive surgery. This study assessed the precision and technical reproducibility of a 3D surgical modelling protocol using free open-source software in mandibular reconstruction with fibula free flaps and surgical guides. Precision was assessed through comparisons of the 3D surgical guide to the sterilized 3D-printed guide, determining accuracy to the millimetre level. Reproducibility was assessed in three surgical cases by volumetric comparison to the millimetre level. For the 3D surgical modelling, a difference of less than 0.1 mm was observed. Almost no deformations (<0.2 mm) were observed post-autoclave sterilization of the 3D-printed surgical guides. In the three surgical cases, the average precision of fibula free flap modelling was between 0.1 mm and 0.4 mm, and the average precision of the complete reconstructed mandible was less than 1 mm. The open-source software protocol demonstrated high accuracy without complications. However, the precision of the surgical case depends on the surgeon’s 3D surgical modelling. Therefore, surgeons need training on the use of this protocol before applying it to surgical cases; this constitutes a limitation. Further studies should address the transfer of expertise.
The surgical gold standard for mandibular bone reconstruction is the fibula free flap . Extensive mandibular bone resection in oncological surgery or traumatic mandibular tissue loss constitutes an indication for reconstructive surgery. The main technical difficulty in such cases is related to the three-dimensional (3D) conformation of the fibula free flap and gaining the most anatomical position possible .
Rapid prototyping (RP) consists of the technology enabling the design, through 3D printing, of objects or models derived from a computed tomography (CT) scan, magnetic resonance imaging (MRI), or optical scan, after a computer-aided 3D modelling stage. RP undoubtedly represents an advance with regard to the various stages of craniomaxillofacial reconstructive surgery planning, but also in the production of prototype surgical instruments, Maxillofacial protheses, and surgical guides . With recent progress made in this technology, it appears to be a particularly pertinent solution with respect to 3D conformation of a fibula free flap in the context of mandibular reconstruction .
Surgeons are now able to finalize the treatment planning using computer-aided 3D modelling and then take the 3D-printed sterilized object into the operating room, thereby improving the precision, control, and duration of the surgical procedure relative to standard techniques . This technology is markedly more precise than that based on pre-moulded patient plaster cast models . However, computer-aided surgical 3D modelling remains the essential preliminary stage for the creation of 3D-printed objects. Since the modelling is complex, it is still mainly implemented by computer engineers and only rarely by surgeons, and is always done using onerous professional software packages. Thus, the main drawbacks are the high cost, most frequently incumbent on the patient, and the long production lead times, usually 10 days to several weeks.
The workflow for this technology in a healthcare organization involves radiologists, surgeons, prosthesis specialists, and the RP department (particularly 3D printing); thus, enhanced organization would enable optimization of the time frame .
Cost reductions and improvements in healthcare quality constitute a surgical challenge. This article reports a technique that would make the technology accessible to all. The technology enables the implementation of surgical 3D modelling by the surgeon through the use of an open-source software protocol and the selection of professional-assistance solutions, with a view to enhancing the timeframe and trimming the costs of the technology, while ensuring the quality of the reconstruction and maintaining patient safety.
This study was designed to evaluate the precision and reproducibility of the open-source protocol using open-source software. The study addressed the precision of the printed surgical guides and analysed three cases of mandibular reconstruction with a fibula free flap using the computer simulations.
Materials and methods
This single-centre study investigated the feasibility and reproducibility of an open-source protocol employing open-source software in the context of mandibular reconstruction. This study involved a single surgeon who specializes in reconstructive head and neck surgery and who is skilled in 3D planning and is the head of student training in this field at a French university.
The open-source software packages used were OsiriX, MeshLab, Netfabb, and Blender. These software programs enabled surgical 3D modelling of the various fibula free flap mandibular reconstructions and the surgical guides.
The RP technology selected is an additive technology – selective laser sintering (SLS) – which uses a thermoplastic in powder form; the thermoplastic material used was polyamide 12 (poly-laurolactam), which is biocompatible .
A 3D printing company was commissioned to implement the models, so that the precision of the surgical modelling using the open-source protocol could be evaluated without a 3D-printing bias. A French company was selected (Sculpteo, Villejuif, France) on the basis of the company’s cost-effectiveness, response times, and the requirement for 3D printing compliant with European quality standards.
Polyamide 12 is a thermoplastic that withstands autoclave steam sterilization without deformation (Belimed, Sausheim, France; 134 °C for 18 min). The 3D object sterilization protocol included cleaning by sonication, followed by rinsing and then drying. The thermoplastic has US Food and Drug Administration (FDA) biocompatibility class VI status.
Technique precision study
In order to validate the precision of the 3D objects printed by Sculpteo and to validate their freedom from deformation post-sterilization, computed tomography (CT) scans were obtained. Comparisons of the scans were made using a fifth open-source software program: CloudCompare. This software enables the comparison of point clouds constituting computer files and generates a colorimetric scale, displaying the deviations between the points in the files compared. These deviations are expressed in millimetres down to values of 10 −2 millimetres.
An ‘ultra-high’ resolution temporal bone CT scan protocol was used to obtain 0.55-mm thick axial cuts, to be as precise as possible (Philips Brilliance 64-channel CT scanner; field of view (FOV) 160 mm, 140 kV, 330 mAs).
The first comparison was made between the stereolithography file (STL) generated from the CT scan of the patient’s mandible and the CT scan of the 3D-printed mandible (printed by Sculpteo) in order to validate the precision of the 3D print in comparison to the CT scan and the computer file. The second comparison was made between the CT scan of the 3D-printed surgical guide post-autoclave sterilization and the 3D modelling of the surgical guide in Blender software, in order to validate the 3D printing process and the absence of polyamide 12 deformation post-sterilization.
Technique reproducibility study
Each surgical guide was approved pre- and post-sterilization for use in the operating room by the surgical team.
Three mandibular fibula free flap reconstruction procedures were performed after the senior surgeon had conducted 3D modelling tailored to each case using the open-source protocol. Any perioperative difficulties with respect to the surgical guide positioning or flap conformation were recorded. The final results with regard to the precision of reconstruction (mm) were derived from a comparison of the patient’s postoperative CT scan and the computer file (STL) of the 3D model using the software CloudCompare, after computerized deletion of the osteosynthesis plates.
The study was authorized by hospital management and complied with the ethical principles defined in the Declaration of Helsinki.