Page 32 - Advanced concepts in orbital wall fractures
P. 32

                                30
Chapter 2
developed, with varying results, that are able to segment the orbit (semi-) automatically using CT scans5,21-23. The clinical applicability of a method should depend on how quick, accurate, reproducible, and versatile it is24.
In this study, manual segmentation of the bony orbit, selected to be the gold standard2,20, is compared to three different methods using iPlan software (version 3.0.5; Brainlab, Feldkirchen, Germany): the automatic segmentation (method A), a semi-automatic method which combines the automatic method with subtraction of a bone (≥ +400 HU) and air (≤ −600 HU) density mask (method SA), and a semi-automatic method minus masks combined with manual adjustments (method SAA). The purpose of this study is to validate these (semi-)automatic segmentation methods for measuring orbital volume based on CT scans of unaffected bony orbits and investigate which method is most suitable for clinical and scientific purposes. The manual segmentation and a newly described delineation of the anterior boundary are first validated to make an accurate comparison possible. To our knowledge, the automatic orbital volume segmentation in this software has not yet been validated for orbital volume segmentation. The software possesses functionalities for preoperative planning and perioperative navigation. The validation of accurate volume analysis serves as a basis for utilising these functionalities in orbital surgery.
Materials and methods
CT data of trauma patients were obtained from the Department of Oral and Maxillofacial Surgery at the Radboud University Nijmegen Medical Centre. From the database of CT scans, a total of 21 orbits, one orbit per scan, was selected. All CT scans were acquired using the standardised trauma protocol (Toshiba Aquillon ONETM): 0.5 mm slice thickness, 0.5 mm slice increment, 100-120 kV, 80-440 mA, 200-220 FOV, 0.656 Pitch, and a 512×512 image matrix. The inclusion criteria were: at least one unaffected bony orbit, no visible orbital pathology, and no blood or other body fluids in the ipsilateral sinuses.





























































































   30   31   32   33   34