Application of 3D printing mirror models and segmentation models in preoperative planning for patellar fractures / 中华创伤骨科杂志

ABSTRACT

Objective:To explore the application of 3D printing mirror models and segmentation models in preoperative planning for patellar fractures.Methods:Retrospectively analyzed were the data of 46 patellar fractures which had been treated at Department of Orthopedics, Liyang City People's Hospital from January 2016 to August 2020 using 3D printing mirror models and segmentation models in preoperative planning. There were 26 males and 20 females, aged from 19 to 79 years (average, 51.5 years). All the fractures belonged to AO type C. According to the design requirements of a solid model, the patients with DICOM CT data of bilateral lower extremities were assigned into the mirror model group (24 cases) while those with DICOM CT data of only a unilateral lower extremity into the segmentation model group (22 cases). In the mirror model group, CT scans of bilateral knee joints and the proximal tibia exceeding 15 cm were required, while in the segmentation model group, CT scans of only the affected knee joint were required. The original DICOM data were extracted and imported into software Mimics 19.0. In the mirror model group, after the "three-level long-short axis control" method was used to judge the symmetry, the original fracture model and the mirror model were created and printed respectively. During the operation, the fracture line drawn by the mirror model was used to find the fracture fragments with a similar shape which were to be assembled and fixated. The fracture fragments in the segmentation model group were divided into independent entities, converted into STL files and printed separately before the bone fragments were assembled and fixated in sequence. X-ray films were taken after operation. Recorded were the simulated operation time, visual analog scale (VAS) of the knee joint at one month after operation, and B?stman scores at one and 6 months after operation.Results:All patients were followed up for 6 to 24 months (mean, 13.6 months). In preoperative planning, only the mirror model was suitable for patients with comminuted fracture with compression and impaction, both models were suitable for fractures without impaction, compression or impaction, and the segmentation model was suitable for patients with bilateral fractures at the same site, poliomyelitis, one limb missing, and previous fracture. In the mirror model group in which the patients were controlled bilaterally using the "three-level long-short axis control" method, there was no significant difference among the 6 sets of data at 3 levels on both sides ( P>0.05). For the segmentation model group and the mirror model group, the time for preoperative simulated operation averaged 11.2 min and 9.2 min respectively, the VAS score at one month after operation was both lower than 3 points, and the B?stman scores at one and 6 months after operation were both larger than 20 points. Conclusions:The mirror image model of the unaffected patella can be used as the fracture recovery model for the affected side, but the symmetry needs to be verified in advance in case of severe degeneration. Both the 3D printed mirror model and the 3D segmentation model are suitable for preoperative planning for patellar fractures as they are complementary. Patients in both groups can obtain good joint function after treatment.