Facial Conformation Meshes for Orthognathic Surgery

Orthognathic surgery for Class III dentofacial deformities has significantly evolved, particularly with advancements in soft tissue prediction technologies. Early methods, such as manual photo manipulation, progressed to two-dimensional (2D) computer-assisted predictions. However, the introduction of three-dimensional (3D) imaging revolutionized the field, enhancing prediction accuracy and enabling comprehensive analysis of facial morphology.

Imaging techniques like Multi-Slice Computed Tomography (MSCT) and Cone Beam Computed Tomography (CBCT) are critical for reconstructing the facial skeleton. MSCT provides high resolution but involves higher radiation exposure, whereas CBCT offers a balance of resolution and lower radiation dose, making it preferable for routine clinical applications. CBCT's ability to capture upright facial soft tissues in a natural head position further enhances its utility in orthognathic planning. Advanced technologies such as stereophotogrammetry and structured light imaging allow detailed capture of facial surfaces. Stereophotogrammetry excels in dynamic facial analysis, while structured light imaging offers rapid acquisition of high-resolution facial data. These techniques facilitate the creation of detailed 3D models essential for accurate surgical planning and prediction.

Soft tissue prediction models rely on mathematical algorithms to simulate responses to skeletal movements. The finite element model (FEM) and mass tensor model (MTM) are widely recognized for their predictive accuracy. MTM balances computational efficiency and precision, making it suitable for clinical applications. Recent innovations, such as probabilistic FEM and machine learning algorithms, further enhance prediction accuracy by integrating patient-specific data.

Generic facial meshes, consisting of indexed vertices, allow for precise superimposition and analysis of 3D facial images. Dense correspondence analysis uses these meshes to provide a comprehensive evaluation of facial surfaces, surpassing the limitations of landmark-based methods. This approach ensures precise prediction of soft tissue changes, contributing to better surgical outcomes.

Modern 3D simulation techniques achieve clinically acceptable prediction accuracies (<2 mm) across various facial regions. These tools not only assist in clinical decision-making but also improve communication with patients by providing photorealistic visualizations of potential surgical outcomes. Recent advancements in 3D simulation and analysis have transformed soft tissue prediction in orthognathic surgery. By integrating dense correspondence methods and patient-specific data, these technologies offer enhanced accuracy and reliability, paving the way for more precise and effective treatment planning for Class III dentofacial deformities. This study aims to evaluate the accuracy of combining a generic mesh-based 3D simulation model and facial esthetic units in predicting soft tissue responses to bimaxillary orthognathic surgery for Class III dentofacial deformities.

4- Aim \& objectives of the Study This study aims to evaluate the accuracy of soft tissue simulation to bimaxillary orthognathic surgery for Class III dentofacial deformities using combination of a conformation mesh-based analysis model and facial esthetic units.

5- Social \& scientific values The clinical relevance of this study lies in the potential to enhance the precision and predictability of treatment outcomes for patients undergoing bimaxillary orthognathic surgery for Class III dentofacial deformities by integrating conformation mesh with facial esthetic units' analysis.

6- Subjects of research:

• Recruitment: No patient recruitment will be conducted, as this study is retrospective and relies solely on existing patient CBCT scans.
• Age N/A 8- Methodology This retrospective study aims to evaluate the accuracy of 3D software in predicting postoperative soft tissue changes following bimaxillary orthognathic surgery for the correction of Class III jaw deformities. The study will utilize pre-existing CBCT images routinely obtained from orthognathic surgery patients at Ain Shams Dental School. The analysis will compare the predicted preoperative soft tissue changes to the actual results observed in six months postoperatively.

Sample Size Calculation:

Based upon the assumption that the 90% is the required prediction accuracy, in light of the results of Resnick CM et al (2016), the computed effect size for the accuracy of linear measurements was found to be (1.05), using alpha (α) level of (5%) and Beta (β) level of (20%) i.e. power = 80%; the study will include 10 subjects. To compensate for 15% for the use of non-parametric tests the final minimum estimated sample size will be 12 subjects. Sample size calculation was performed using G*Power Version 3.1.9.2.

CBCT scans of Class III patients who underwent maxillary advancement with minimal or no impaction and mandibular setbacks will be identified. To quantify the actual surgical movements, voxel-based superimposition of the pre- and postoperative CBCT images will be conducted using ONDEMAND 3D software. Surgical movements will then be calculated through direct slice landmarking. These movements will inform the soft tissue prediction planning, which will subsequently be compared with the actual postoperative changes.

Soft tissue surface models for both the simulation and actual postoperative outcomes will be generated and exported as STL files. Superimposition of the predicted and postoperative STL models will be performed, and generic meshes will be conformed. Using VR Mesh software, the facial meshes will be segmented into predefined anatomical regions. Distances between corresponding vertices on the predicted and postoperative models will be displayed as a color scale in millimeters. Analyses will be conducted separately for the mediolateral (x), vertical (y), and anteroposterior (z) dimensions using custom-developed in-house software. Color-coded distance maps will be generated to illustrate the magnitude of prediction inaccuracies in each anatomical region. Each dimension will be analyzed to calculate the minimum, maximum, mean, standard deviation (SD), absolute maximum, and mean SD for 90% of the points in each region.

Dimensional inaccuracies will be represented visually, with corresponding points showing rightward, upward, or anterior changes displayed in red, while leftward, downward, or posterior changes will appear in blue. This approach will enable a detailed assessment of prediction accuracy across the three spatial dimensions.

9- Risks to study participants No Risks as no participants will be involved, the study is retrospective utilizing existing data. No direct patient involvement will be required, and all data was fully anonymized before analysis.

10- Benefits to study participants/ community This study benefits lies in the potential to enhance the precision and predictability of treatment outcomes for patients undergoing bimaxillary orthognathic surgery for Class III dentofacial deformities by integrating conformation mesh with facial esthetic units' analysis.

11- Privacy, Record keeping \& confidentiality In this retrospective study Patient data will be anonymized using DICOM anonymizer software and assigned unique patient identification numbers. The anonymized data will be securely stored in an Excel sheet for easy retrieval of individual patient information. All data will be maintained on a limited-access, password-protected computer with the principal investigator. If analysis requires transferring images to another computer, a hardware-encrypted removable storage device will be used, and the data will be deleted from the device immediately after transfer. Access to the research data will be restricted to members of the research team only.

12- Consent form: A waiver for individual informed consent from the Institutional Review Board is considered, as the study is retrospective based on existing patient data.