The Combined Application of Database and Three-Dimensional Image Registration Technology in the Restoration of Total Nose Defect.

Abstract

The purpose of this study is to present a virtual planning protocol based on the database and three-dimensional (3D) image registration technology for the restoration of the total nasal defect, and evaluate its feasibility and clinical efficacy. Patients were scanned with a FaceScan to obtain the 3D facial model which was stored in an Extensible Neuroimaging Archive Toolkit (XNAT) database. Personalized search and similarity evaluation were performed in the database to find a normal 3D facial model with the highest similarity to the patient's. Then, the 2 models were imported into the dedicated software for 3D image registration to get the 3D model of the nasal defect part and the preoperative planning 3D facial model of virtual restoration. Subsequently, the dimensionality reduction algorithm was conducted to transform the 3D model of the nasal defect to a 2D flatten one for determining the scope of the forehead flap during surgery. Four weeks after the insetting surgery of forehead flap pedicle, the postoperative 3D facial model was gained. At last, the clinical efficacy was evaluated by comparing the difference between the preoperative planning and postoperative 3D facial model. The nasal shapes of the patients were good after the operation, and the results revealed that the maximum error was ranging from 3.12 to 4.07 mm with the mean error from 0.92 to 1.04 mm. The database and 3D image registration technology provide a new approach for precisely determining the scope of total nasal defect and the forehead flap, which may be used for reference in the accurate restoration of other facial soft tissue defects.