Augmented reality hologram combined with pre-bent distractor enhanced the accuracy of distraction vector transfer in maxillary distraction osteogenesis, a study based on 3D printed phantoms.

Background: Vector control is a significant concern in maxillary distraction osteogenesis (DO). Distraction vector planning on the patient's 3D-printed skull phantom is more intuitive for surgeons and cost-efficient than virtual surgical planning. However, the accuracy of transferring the planned vector to intraoperative (vector transfer) according to the shape of the pre-bent footplate alone is relatively limited. The application of augmented reality (AR) in surgical navigation has been studied for years. However, few studies have focused on its role in maxillary DO vector transfer. This study aimed to evaluate the accuracy of AR surgical navigation combined with the pre-bent distractor in vector transfer by comparing it with the pre-bent distractor alone.

Methods: Ten patients with maxillary hypoplasia were enrolled with consent, and three identical 3D-printed skull phantoms were manufactured based on per patient's corresponding pre-operative CT data. Among these, one phantom was for pre-operative planning ( = 10), while and the other two were for the AR+Pre-bending group ( = 10) and the Pre-bending group ( = 10) for the experimental surgery, respectively. In the Pre-bending group, the distraction vector was solely determined by matching the shape of footplates and maxillary surface. In the AR+Pre-bending group, the distractors were first confirmed to have no deformation. Then AR surgical navigation was applied to check and adjust the vector in addition to the steps as in the Pre-bending Group.

Results: For the angular deviation of the distraction vector, the AR+Pre-bending group was significantly smaller than the Pre-bending group in spatial ( < 0.001), x-y plane ( = 0.002), and y-z plane ( < 0.001), and there were no significant differences in the x-z plane ( = 0.221). The AR+Pre-bending group was more accurate in deviations of the Euclidean distance ( = 0.004) and the y-axis ( = 0.011). In addition, the AR+Pre-bending group was more accurate for the distraction result.

Conclusions: In this study based on 3D printed skull phantoms, the AR surgical navigation combined with the pre-bent distractor enhanced the accuracy of vector transfer in maxillary DO, compared with the pre-bending technique alone.