Page 71 - Advanced concepts in orbital wall fractures

Methods
This study design was extensively introduced and discussed in
earlier studies5,7,8. In the first group, one surgeon (LD) performed the reconstructions without three-dimensional (3D) virtual diagnostics and
preoperative planning by using the transconjunctival approach. The
surgeon was allowed to have a look at the preoperative scan. The surgeon
was unaware of which cadaver he was operating on. The orbital titanium
mesh plates were put in place and fixated with one bone screw in the
inferior orbital rim. After completion, all cadavers were scanned, the
mesh plates were removed and the drill holes were covered with DuraLay
(Reliance Dental Mfg. Co, Worth, IL, USA). In the second group, the same
surgeon repeated the process several months later, but now with the use C of 3D diagnostics and a virtual preoperative plan of the reconstruction. 4 The surgeon could consult the planned position of the implant as a virtual
guide before and during surgery on a computer screen in the operating theatre7. To check intrasurgeon and intersurgeon repeatability for both groups, the reconstructions were repeated by two surgeons (LD and PG).
iPlan software was used to evaluate the accuracy of the implant position. The positioned implant was segmented (threshold Hounsfield units >1200) and transferred as a STL file. The Orbital Implant Positioning Frame (OIPF) superimposed the postoperative implant onto the planned preoperative implant position for every reconstruction20. The OIPF calculated differences in translation and rotations (pitch, yaw, and roll) in relation to the ideal planned position using the iterative closest point approach (Fig. 1). Translation was measured as total displacement, resulting in the Euclidean distance of the translation in the x, y, and z direction.
Statistical analysis
Statistical data analysis was performed using SPSS Statistics (version 22.0; IBM Corp., Armonk, NY, USA). For all results the pitch, yaw, roll, and translation of the ideal planned position was compared to the postoperative actual position. This was performed for both groups (with and without preoperative planning). Intersurgeon and intrasurgeon variability was calculated for both surgeons using the intraclass correlation
Preoperative planning
69