Prenatal craniofacial abnormalities: from ultrasonography to three‐dimensional printed models

Three-dimensional (3D) ultrasound has improved significantly the prenatal diagnosis of cleft lip/palate and other craniofacial deformities, facilitating the provision of preoperative information that was previously not available. Recent improvements in imaging systems and advances in prenatal 3D ultrasound have allowed further optimization of the diagnosis of such abnormalities, providing parents and healthcare practitioners with more comprehensive morphologic information1. This technology contributes to helping parents understand the pathology of their fetus and facilitates communication between parents and the surgical team. Surface rendering of the fetal face is, as such, a valuable prenatal tool. Nevertheless, current 3D image reconstructions can be displayed only in two dimensions (2D), potentially representing a conceptual limit for the complete understanding of the fetal morphology, through a lack of perception of depth of field. To overcome this limitation, a 3D printed haptic model may be used. This technology allows both practitioners and parents to develop a better understanding of the anatomy and treatment procedure, as they can interact directly with the model2. Data for 3D printed medical models are traditionally obtained by computed tomography or magnetic resonance imaging, using Data Imaging and Communications in Medicine (DICOM) software. In 2017, we reported on the first printed model obtained using data from a 3D sonographic oropalatal view2. Since then, manufacturers have developed an output compatible with a numerical format recognized by a 3D printer device, and several 3D printed fetal models have been reported3-5. Here, we present 3D printed models demonstrating fetal craniofacial abnormalities obtained using data from prenatal surface-rendered sonographic views. Ultrasound volumes were acquired using a Voluson E8 machine (GE Healthcare, Zipf, Austria) with conversion of the .VOL file to .STL (standard tessellation language), or using a Voluson E10 machine (GE Healthcare) with .STL output. We used RAB 4-8, RM 6C or RAB 6 transabdominal probes. The acquisition protocol is based on multiplanar reconstructions in the three planes associated with surface reconstructions of the area of interest. The .STL file is then corrected using software such as Netfabb® (Autodesk, Inc., San Rafael, CA, USA) before printing, in order to correct errors and trim the margins of the model to obtain a closed .STL file (an informatic model consisting of a fully closed mesh). The file is then loaded in Up Studio® software (Beijing Tiertime Technology Co. Ltd., Beijing, China), which allows communication with the printer and defining the print settings and scaffolding required for printing. The print settings are changed according to the quality of the template and the desired print time. For example, in a previous study, an acrylonitrile-butadiene-styrene (ABS) 3D printed model of a fetal face with cleft lip and palate was created using 0.15-mm layer thickness (the thinnest slicing available for the printer used) for a total duration of 6 h and 52 min, for an ABS weight of 72.1 g2. Printing is then done using a ‘low-cost’ UPplus2® printer (Beijing Tiertime Technology Co. Ltd.). This 3D printer is a rapid prototyping machine based on a fused deposition modeling technique that uses ABS or polylactic acid as the raw material. We chose to use ABS as it seems to provide better definition in printed models. Plastic ABS models printed using the low-cost 3D UPplus2 printer provide dimensional accuracy comparable to that of other well-established high-cost professional rapid prototyping technologies2. As the printer cannot print midair, some objects require simultaneous printing of support areas. The 3D model must therefore be cleaned of any scaffolding using a scaffolding clip. 3D printed ABS haptic models of the faces of three fetuses with a facial abnormality were created (Figure 1). Figure 1a shows the model of a 32-week fetus with Crouzon syndrome, which had a printing duration of 6 h and 31 min, for an ABS weight of 65.9 g. Figure 1b shows the model of a 24-week fetus with cleft lip and palate, which had a printing duration of 6 h and 52 min, for an ABS weight of 72.1 g. Figure 1c shows the model of a 24-week fetus with proboscis nasalis lateralis, which had a printing duration of 5 h and 55 min, for an ABS weight of 55.4 g. We considered several points with regards to whether this new technology could be of added value in daily clinical practice. The models can be used as a tool to optimize information given to parents in, for example, cases of a fetal complex craniofacial malformation. The spatial representation of abnormal morphology can be difficult for an untrained person to understand from ultrasound images displayed in 2D. Therefore, a 3D object that can be manipulated and mobilized in space can facilitate the understanding of such a malformation2. 3D objects may also help guide the surgical team in cases of a complex malformation, such as cardiac or spinal malformation, which are characterized by spatial complexity. 3D mapping helps the surgeon to optimize the planning of the surgical procedure2,4. Finally, haptic models may be a medium of choice for teaching2,5, as they can be manipulated and annotated. As part of these counseling and teaching activities, the use of a set of standard 3D printed models showing typical pathologies may be of value. For example, in a clinical context, it is not always possible to perform 3D printing due to technical or organizational difficulties. A set of standard pathological models could therefore be of use in these instances. 3D printing using ultrasound data is an emerging tool both in clinical practice and in teaching. However, many elements must be evaluated and improved to clarify its role in the management of pregnancies with a fetal malformation. Technical characteristics, such as the type of material used and printing time, still need to be optimized. Finally, the parents' view of an anomaly becoming palpable prenatally has ethical considerations that remain to be evaluated.