Comparison of Clinical Outcomes Between Fixed Orthodontic Appliances and Clear Aligners Following Orthognathic Surgery: A Systematic Review and Narrative Meta-Analysis

김준하 경희대학교 대학원 2026 국내박사

Background: Postoperative orthodontic treatment is essential for stabilizing skeletal and occlusal outcomes following orthognathic surgery. Although fixed appliances have traditionally been the standard of care, the use of clear aligners has increased due to their advantages in removability, aesthetics, and oral hygiene maintenance. In adult surgical patients, early postoperative discomfort, soft-tissue recovery, and plaque control are particularly important factors influencing treatment adherence and overall outcomes. However, comparative evidence regarding the effectiveness of clear aligners versus fixed appliances in the postoperative phase remains limited and inconsistent. Objectives: This study aimed to systematically evaluate and compare the clinical effectiveness, periodontal outcomes, skeletal and occlusal stability, postoperative recovery indicators, and patient-reported outcomes of clear aligner therapy and fixed orthodontic appliances in adults undergoing postoperative orthodontic treatment after orthognathic surgery. Methods: A systematic review was conducted in accordance with the PRISMA 2020 guideline. Electronic searches were performed in PubMed, Embase, Cochrane CENTRAL, Web of Science, and Scopus for studies published between January 2015 and August 2025. Comparative studies involving adults aged 16 years or older who underwent single- or double-jaw orthognathic surgery followed by postoperative orthodontic treatment with either clear aligners or fixed appliances were included. Risk of bias was assessed using RoB 2.0 for randomized trials and ROBINS-I for observational studies, and the certainty of evidence was evaluated using GRADE. Due to heterogeneity in study designs and outcome measures, a narrative synthesis was conducted. Results: A total of seven studies, including one randomized controlled trial and six retrospective comparative studies, met the inclusion criteria. Clear aligners demonstrated consistently favorable periodontal outcomes, such as lower plaque index, reduced probing depth, decreased bleeding on probing, and less root resorption. Patient-reported outcomes also favored clear aligners, showing improvements in oral health-related quality of life, orthognathic-specific quality of life, esthetic and functional satisfaction, and lower early postoperative pain. Early soft-tissue swelling was reduced in clear aligner users, although differences diminished after three to four weeks. Skeletal and occlusal stability, treatment duration, operative time, hospital stay, and complication rates showed no meaningful differences between clear aligners and fixed appliances. The certainty of evidence ranged from low to moderate. Conclusion: Clear aligner therapy provides comparable skeletal and occlusal stability to fixed appliances while offering notable advantages in periodontal health, soft-tissue recovery, and patient experience in adult patients following orthognathic surgery. Clear aligners represent a clinically acceptable and patient-centered alternative for postoperative orthodontic treatment; however, larger, well-designed randomized controlled trials with long-term follow-up are required to strengthen the current evidence base and guide optimal clinical decision-making.

Impulse-based Optimal Dental Occlusion Simulation for Orthognathic Surgery

Seunghye Jeong DGIST 2021 국내석사

This paper deals with the three-dimensional dental occlusion model required when planning Orthognathic surgery using Computer-Aided Surgical Simulation (CASS). Orthognathic surgery is a treatment for severe malocclusion, and the optimal dental occlusion that can be obtained immediately after the surgery is called optimal occlusion. In CASS, the cutting and movement of the jaw bone is planned so that the existing occlusion of the CT-based 3D skull model is matches the optimal occlusion model. The stone model of the upper and lower dental is fabricated to obtain the 3D optimal occlusion model. The stone models are moved manually ant the most stable occlusal position is determined using tactile and visual feedback. This process is time consuming and material consuming and does not quantitatively check the occlusal contact area. And these days, with the advancement of intraoral scanners, a 3D dental model is immediately acquired. Therefore, it is necessary to automatically determine optimal occlusion in 3D. There are previous studies to determine optimal surgical occlusion automatically using 3D dental models, but they do not guarantee results in cases of occlusal interference or the lack of teeth. However, for treatment purposes, the surgery is often performed under these situations intentionally. Therefore, in this paper, we propose an impulse-based occlusion optimization method to automatically determine optimal occlusion in 3D including relatively abnormal teeth. In addition, it includes Occlusal Equilibration, an irreversible treatment for severe occlusal interference by premature contact. Simulation results are verified by comparison with expert results. 본 논문에서는 Computer-Aided Surgical Simulation 을 이용하여 양악 수술을 계획 할 때 필요한 3 차원 치아 교합 모델에 관하여 연구합니다. 양악 수술은 중증 부정 교합에 대한 치료법으로 수술 직후에 얻을 수 있는 최적의 치아교합을 최적 교합이라고 합니다. CASS 에서 CT 기반의 3 차원 두개골 모델의 기존 교합이 최적 교합 모델과 일치하도록 턱 뼈의 절단 및 이동을 계획합니다. 현재까지도 3 차원 최적 교합을 얻기 위해서 상악과 하악의 석고 모형을 제작합니다. 석고 모델을 수동적으로 움직이며 촉각적, 시각적 피드백을 사용하여 가장 안정적인 교합 위치를 결정합니다. 이는 많은 시간과 재료가 소요되며 교합 접촉 면적을 정량적으로 확인할 수 없습니다. 또한 요즘에는 구강내 스캐너의 발전으로 직접 3 차원 치아 모델을 바로 획득할 수 있습니다. 따라서 3 차원에서 최적교합을 자동으로 결정하는 연구가 필요합니다. 이전 연구들은 발치로 인한 빈공간이나 조기접촉과 같은 교합간섭이 존재하는 경우를 보장하지 않습니다. 그러나 치료를 위해 의도적으로 이러한 상황에서 수술이 수행되곤 합니다. 따라서 본 논문에서는 비교적 비정상적인 치아를 포함하여 3 차원에서 자동으로 최적 교합을 결정하기 위한 임펄스 기반의 교합 최적화 방법을 제안합니다. 또한 조기접촉으로 인한 심각한 교합간섭의 비가역적 치료법인 OE 단계를 포함합니다. 시뮬레이션 결과는 전문가 결과와 비교하여 검증하였다.

Development of splintless orthognathic surgery methods : Comparison between maxillary template-based surgery and robot arm with navigation assisted surgery

한정준 서울대학교 대학원 2018 국내박사

Background Le Fort I osteotomy is widely used in orthognathic surgery to reposition the maxilla to a balanced and optimal position. Correct repositioning of the maxilla to the planned position is essential for achieving a successful surgical outcome in orthognathic surgery. In order to transfer a surgical plan to the actual operation, most surgeons use intermediate splints. This involves several error-prone procedures, including making a cast model, face bow transfer, model surgery, and fabrication of acrylic splints. In addition to these errors during preoperative preparation, malpositioning of the maxilla can also occur during surgery due to the perioperative condylar displacement. Thus, a template-based orthognathic surgery and a robot arm with navigation assisted orthognathic surgery were designed to replace the conventional maxillary repositioning methods that involves the use of intermediate splint. First, we developed the template-based orthognathic surgery and evaluated the accuracy of the positioning of the CAD/CAM based templates on dry skull, because its precise positioning is prerequisite for successful outcome. For the robot arm with navigation assisted surgery system, we evaluated the validity and reliability of the system for various surgical plans using phantom skulls. Materials and methods I. Using the computed tomography (CT) data, CAD/CAM based maxillary surgical templates for orthognathic surgery were developed. Maxillary surgical templates consisted of 2 templates: the osteotomy guide template and the repositioning guide template. The osteotomy guide template had several specialized structures for exact and passive positioning according to the simulation surgery and guide the osteotomy and removal of bony interference. The repositioning guide template guided the maxilla to the planned position using wedge-shaped male and female structures. To determine the accuracy of the template positioning, deviation and fitness of the template were evaluated after positioning of the template on the dry skull. II. The robot arm with navigation assisted orthognathic surgery was developed and consisted of a robot arm with six degrees of freedom, an image-guided navigation system including an optical tracking camera and a tracking tool, a display and a computer. The end-effector of the robot arm was designed to connect the robot arm to the surgical splint and to enable identification and removal of any bony interference. The position of robot arm was tracked in real time, and the robot arm and osteotomized maxilla were moved to the planned position using image-guided navigation. During the fixation, the osteotomized maxilla was held by the end-effector of the robot arm. In order to evaluate the accuracy of robot arm with navigation assisted orthognathic surgery, experiments were conducted using 12 full skull models. A total of 12 experiments involving four different surgical plans were conducted, with each surgical plan being implemented three times. Using intraoperative navigation and postoperative CT data, deviations between the planned and actual positions of the maxilla were calculated. Results I. Regarding the deviation of the CAD/CAM based surgical templates, the mean absolute deviation was 0.41 ± 0.30 mm in medial-lateral direction (p < 0.001), 0.55 ± 0.59 mm in anterior-posterior direction (p < 0.001) and 0.69 ± 0.59 mm in superior-inferior direction (p < 0.001). The deviation was statistically significant in all directions. The mean root mean square deviation (RMSD) between the planned and the actual position of the template was 1.07 ± 0.76 mm. With respect to the fitness of the template, the mean distance between the inner surface of the template and the underlying bone was 0.76 ± 0.24 mm (range, 0 to 2.38 mm). II. All experiments conducted using the developed robot arm with navigation assisted surgery system were successful. In the assessment of maxillary repositioning during intraoperative navigation, the mean absolute deviations were 0.10 ± 0.12 mm, 0.08 ± 0.07 mm, and 0.10 ± 0.10 mm in the medio-lateral, antero-posterior, and supero-inferior directions, respectively. The mean RMSD was 0.19 ± 0.15 mm. During the stabilization of the maxilla with miniplates, the maxillary segment showed slight deviations of 0.22 ± 0.19 mm, 0.18 ± 0.12 mm, 0.18 ± 0.12 mm, and 0.37 ± 0.20 mm in medio-lateral direction (p < 0.001), antero-posterior direction (p = 0.303), supero-inferior direction (p = 0.426), and in three-dimensional space, respectively, as compared to the maxillary position before stabilization. The final surgical outcome assessed using postoperative CT data showed that the mean deviations between the planned and actual positions were 0.42 ± 0.35 mm, 0.37 ± 0.25 mm, and 0.38 ± 0.32 mm in the medio-lateral (p = 0.169), antero-posterior (p = 0.006), and supero-inferior directions (p = 0.003), respectively, and the mean RMSD was 0.79 ± 0.35 mm. The error in superimposition for the comparison between the preoperative planned and the actual postoperative 3D model was 0.05 ± 0.01 mm. Surgical accuracy was similar in all regions of the maxilla. Conclusion This study suggests that various surface-based CAD/CAM templates can be positioned based on prior simulation without the help of occlusal splints, and can serve as an alternative to traditional methods that involve the use of intermediate splints for the transfer of surgical plans to the actual operation. However, these templates have some limitations with regard to their universal clinical applicability. The robot arm with navigation assisted orthognathic surgery system allows us to overcome the limitations of template-based orthognathic surgery; the maxilla can be repositioned and stabilized successfully, with accuracy comparable to or better that that of other methods of surgical plan transfer.

악교정환자의 수술전후 발음변화에 관한 음향학적 특성

전경숙 전북대학교 대학원 2007 국내석사

It is reported that the orthognathic patients suffer from not only aesthetic problems but also resonance disorder and articulation disorder because of the abnormality of the oral cavity. These disorders have an influence on the patients' communication and they are usually required to be corrected by orthognathic surgery. Speech of the orthognatic patients is affected by the change of the oral cavity capacity and structures after surgery. This study was designed to investigate the resonance of nasality and the intelligibility of speech for acoustic characteristics of speech for pre and post orthognatic surgery patients. 8 orthognathic patients were participated in the study. In order to analyze the nasality of word containing Korean consonants, Korean consonants and frequency and intensity of the fricative /s-/, Nasometer and CSL(computerized Speech Lab) were used. The results were as follows: 1. The nasality of post orthognathic surgery patients decreased in spontaneous speech. There was a significant difference in the nasality for all words between pre and post orthognatic surgery patients. 2. The nasality of each Korean consonant phoneme of post orthognathic surgery patients decreased. There was also a significant difference of the nasality for each Korean consonant phoneme between pre and post orthognatic surgery patients. 3. The decreased nasality for Korean consonant phonemes showed in plosives, affricatives, fricatives, liquids, and nasals after surgery. But the significant difference showed only in plosives and fricatives. 4. Frequency and intensity for the fricative /s/ of post orthognathic patients increased. The purpose of this study was to demonstrate the characteristics of speech for pre and post orthognathic surgery patients as well as to provide clinicians with valuable data based on the acoustic analysis of the orthognathic patients' speech.

Development of Soft Tissue Prediction Model after Orthognathic Surgery by Applying Artificial Intelligence Based on TabNet Deep Neural Network

박지애 서울대학교 대학원 2024 국내박사

연구 목적: 수술 결과를 예측하는 것은 임상 교정 의사가 수술 후 연조직 변화를 시각화하여 치료 계획을 수립하고 만족스러운 치료 결과를 달성하기 위해 필수적이다. 수십 년 동안 턱교정 수술 후 연조직 변화를 예측하기 위한 많은 시도가 있었다. 그러나 기존의 최소제곱 (Ordinary least squares, OLS) 분석을 포함한 이러한 방법의 예측 결과는 치아, 골격 및 연조직 간의 상호 상관관계 때문에 실제 결과와 차이를 나타내었다. 한편 부분최소제곱 (Partial least squares, PLS) 방법은 서로 상관관계가 높은 변수가 많을 때 사용할 수 있다. 몇몇 연구에서는 턱교정 수술 후 연조직 변화 예측에 PLS 방법의 적용을 제안하였으며, 수술 후 연조직 변화 예측 시 PLS 방법이 OLS 방법보다 더 정확하게 예측할 수 있다고 밝혔다. 최근 치과 영역에 인공지능 (Artificial Intelligence, AI)을 적용하려는 시도가 크게 증가하고 있다. TabNet은 최근 개발된 심층 신경망 (Deep neural network, DNN) 구조로서, TabNet 알고리즘은 다양한 입력 및 출력 변수를 효과적으로 처리하여 수술 예측에 사용될 수 있을 것이다. 본 연구의 목적은 TabNet 심층 신경망을 기반으로 한 AI를 적용하여 턱교정 수술 후 연조직 예측 모형을 개발하는 것이다. 그리고 인공지능을 사용하여 구축한 모형을 PLS 방법을 적용하여 구축한 예측 모형과 비교하는 것이다. 재료 및 방법: 2002년 1월부터 2022년 12월까지 턱교정 수술을 받은 환자 705명의 수술 전 및 치료 후 측모 두부계측방사선사진을 연구 대상으로 하였다. 다변량 PLS방법과 TabNet 심층 신경망 AI 알고리즘을 적용하여 턱교정 수술 후 연조직 변화 예측 모형을 개발하였다. 예측 변수는 254개, 반응 변수는 64개가 포함되었다. 두 예측 모형의 정확도를 비교하기 위해 t-검정이 시행되었으며, 수술 예측 모형의 성능에 영향을 미치는 요인을 결정하기 위해 다중 선형 회귀 분석이 시행되었다. 수술 후 연조직 반응의 변동성과 패턴이 예측모델의 정확도에 미치는 영향을 평가하기 위해 신뢰 타원을 도해하였다. 결과: 최적의 AI 기반 예측 모형은 10000회의 훈련이 시행되었을 때 도출되었다. PLS 기반 모형과 AI 기반 모형 간의 절대 오차의 차이는 유의하지 않았다. 하악과 턱에 위치한 6개의 계측점에서 AI가 더 나은 예측을 보인 반면, 두개골 기저부와 상악을 중심으로 한 16개의 계측점에서 PLS가 더 정확했다. 반면 다른 계측점에서의 예측 결과는 두 모형 간에 큰 차이가 없었다. 환자의 연령에 따른 절대 오차의 차이는 발견되지 않았다. 그러나 예측 오차는 수술 후 경과시간(연간 0.07 mm)에 비례하여 증가하였다. 또한 여성, III급 또는 I급 부정교합 대상자, 수평 방향의 수술적 재위치, 상악의 계측점에서 각각 남성, II급 부정교합 대상자, 수직 방향의 수술적 재위치, 하악의 계측점보다 더 정확한 예측 결과를 보였다. 결론: PLS 방법과 TabNet AI 알고리즘은 수많은 변수를 통합한 모델 구축이 가능하기 때문에 악교정 수술 후 연조직 변화를 예측하는 데 사용하기에 적합해 보인다. TabNet AI 알고리즘은 특히 menton 이하의 가변성이 높은 일부 계측점을 우수하게 예측하는 것으로 나타났다. 주요어: 연조직 예측, 악교정수술, 부분최소제곱 방법, 인공지능 학번: 2020-33464 [Introduction] Predicting surgical outcomes is imperative for clinical orthodontists to formulate treatment plans and achieve satisfying treatment results by visualizing post- operative soft tissue change. There have been a lot of attempts for predicting soft tissue change after orthognathic surgery during several decades. However, the prediction results from these methods including the ordinary least squares (OLS) analysis were far from the real outcomes because of the intercorrelation between teeth, skeleton and soft tissue. The partial least squares (PLS) method can be employed when there are a lot of variables highly correlated with each other. Some studies suggested the application of the PLS method on predicting soft tissue change after orthognathic surgery. It was revealed that the PLS method can predict post-operative soft tissue change more accurately than the OLS method. The attempts to adopting artificial intelligence on dentistry have been considerably increasing. TabNet was a recently developed deep neural network architecture for processing tabular data. In this regard, TabNet algorithm may be useful for the surgery prediction by handling a number of input and output variables effectively. This study aimed to develop soft tissue prediction model after orthognathic surgery by applying artificial intelligence based on TabNet deep neural network and compare the AI-based model with the PLS-based model. [Materials and methods] Preoperative and posttreatment lateral cephalograms from 705 patients who had undergone orthognathic surgery from January 2002 to December 2022 were included in this study. On a total of 1410 images, 46 skeletal and 32 soft-tissue landmarks were identified. Soft tissue prediction models after orthognathic surgery were built by applying the multivariate PLS regression and a deep learning method based on the TabNet deep neural network (AI), including 254 predictor and 64 response variables. In prediction accuracy, two prediction models were compared with each other by conducting the t-tests. The multiple linear regression analysis was performed to determine factors influencing the prediction accuracy of the soft tissue prediction models after orthognathic surgery. To evaluate the effect of variability and pattern of soft tissue response after surgery on the accuracy of prediction models, 95% confidence ellipses were depicted. [Results] The optimal AI-based prediction model has resulted from 10000 times training. The difference of absolute error between the PLS-based model and the AI-based model was not significant. AI showed better prediction in 6 landmarks located in mandible and chin, whereas PLS was more accurate in 16 landmarks, mainly from cranial base and maxilla. The prediction results in the other landmarks were not significantly different between two models. Additionally, the difference of absolute error was not found according to patients’ age. However, the prediction error increased in proportion to the elapsed time after surgery (0.07 mm per year). Females, subjects with Class III or Class I malocclusion, horizontal component of surgical repositioning, and landmarks on the maxilla showed more accurate prediction results than males, subjects with Class II malocclusion, vertical component of surgical repositioning, and landmarks on the mandible, respectively. [Conclusions] Since the PLS method and the TabNet AI algorithm are capable of building models that incorporate numerous variables, these algorithms seem satisfactory for use in predicting soft tissue change after orthognathic surgery. The TabNet AI algorithm was outperformed in predicting some landmarks with high variability, especially below menton. Key Words: Soft tissue prediction; Orthognathic surgery; Partial least squares; Artificial Intelligence Student Number: 2020-33464

Comparison of transverse dental axis changes in skeletal Class III with asymmetry treated by preorthodontic orthognathic surgery and conventional surgery : 비대칭을 동반한 골격성 III급의 술전교정 유무에 따른 횡적 치축 변화

Song, Han-Sol 연세대학교 일반대학원 2016 국내석사

본 연구에서는 비대칭을 동반한 골격성 III급의 수술 전과 후의 횡적 치축 변화를 3차원 CT를 통해 분석하고, 선수술과 통상적인 악교정수술 간에 변화양상을 비교하였다. 총 29명의 골격성 III급 부정교합 환자 (남 : 15명, 여 : 14명) 가 선정되었으며, 이 중 16명 (남 : 10명, 여 : 6명, 평균나이 21.8 ± 2.2세) 은 통상적인 악교정수술로, 13명 (남 : 5명, 여 : 8명, 평균나이 21.0 ± 1.7세) 은 선수술로 치료받았다. 상하악 악교정수술은 상악은 Le Fort I osteotomy, 하악은 bilateral intraoral vertical ramus osteotomy (B-IVRO) 로 진행하였으며, 치료 시작 전 (T0), 수술 1달 전 (T1), 수술 3일 후 (T2), 수술 1년 후 (T3) 3차원 CT를 촬영하여 골격 및 치아계측 항목의 횡적 변화량을 분석한 결과 다음과 같은 결론을 얻었다. 1.통상적 수술군에서 술전교정에 의해 (T0-T1) 상악 제1대구치와 견치는 직립되었으나, 술전교정을 했음에도 불구하고 수술직전 (T1) 통상적 수술군과 선수술군 간 상하악 치아의 협설측 치축경사에는 유의미한 차이가 없었다. 2.수술 후 1년 동안 (T2-T3) 선수술군에서 좌우측 상악 제1대구치와 비편위측 하악 견치는 설측경사를 보였다. 3.수술 1년 후 (T3), 술전교정 유무에 따른 상하악 견치와 제1대구치의 협설측 치축 경사의 유의할 만한 차이는 없었으나, 선수술군의 편위측-비편위측 제1대구치의 치축 경사는 차이가 다소 잔존하였다. 4.통상적인 수술군의 술전교정 기간은 (T0-T1) 평균 12개월이었으나, 치료 전과 수술 1년 후의 (T0-T3) 상하악 견치와 제1대구치의 횡적인 치축변화는 통상적인 수술군과 선수술군 간 유의한 차이가 없었다. 안모비대칭을 동반한 골격성 III급에서, 선수술은 통상적인 악교정수술과 횡적인 치축 분석에서 비슷한 치료결과를 보이며, 치료기간의 단축 및 즉각적인 심미개선의 측면에서 효율적인 치료로 볼 수 있다. 하지만 편위-비편위측간의 협설 치축 경사 차이가 매우 심한 경우, 보다 안정적인 구치부 교합관계를 위하여 술전교정으로 decompensation을 어느 정도 시행한 후 악교정수술을 시행하는 것을 고려해볼 수 있겠다. The aim of this study was to evaluate, using 3D-CT, transverse dental axis changes in skeletal Class III patients with facial asymmetry who received 2-jaw surgery, and compare the changes between two different orthognathic surgery protocols; conventional surgery and preorthodontic orthognathic surgery. Total 29 skeletal Class III patients (15 men, 14 women) with asymmetry (menton deviation > 4 mm) were selected. Sixteen patients (CS group, 10 men, 6 women, mean age 21.8 ± 2.2 yrs) received conventional orthognathic surgery, and thirteen patients (POGS group, 5 men, 8 women, mean age 21.0 ± 1.7 yrs) received preorthodontic orthognathic surgery. Orthognathic surgery was proceeded with maxillary Le Fort I osteotomy and mandibular bilateral intraoral vertical ramus osteotomy (B-IVRO). Facial 3D-CT was taken before treatment (T0), 1 month before surgery (T1), 3 days after surgery (T2) and 1 year after surgery (T3). Skeletal and dental variables were measured and compared the dental axis changes between CS and POGS groups. The results are as followings, 1. By pre-surgical orthodontic treatment in CS group, upper 1st molar and canine had shown tendency of uprighting, but there was no significant difference between CS and POGS group at 1 month before surgery (T1). 2. In POGS group, lower canine of non-deviated side and upper 1st molar of both side inclined lingually by post-surgical orthodontic treatment (T2-T3). 3. One year after the surgery (T3), skeletal and dental measurements were similar between CS and POGS groups. However, there was significant difference of buccolingual inclination of maxillary and mandibular molars between deviated and non-deviated side in POGS group. 4.Although in CS group the pre-surgical orthodontic treatment (T0-T1) was done for about 12 months, total changes of buccolingual inclination of canine and molar showed no significant difference between CS and POGS groups (T0-T3). Preorthodontic orthognathic surgery can be an efficient method for skeletal Class III patients with facial asymmetry in the aspect of shorter treatment time and immediate improvement of facial esthetics, and lead to almost the same treatment outcome comparing to conventional surgery. However, if there is severe difference of buccolingual inclination between deviated and non-deviated posterior teeth, conventional surgery with dental decompensation prior to orthognathic surgery would be a better surgical option for a relatively stable treatment outcome.

Clinical Analysis of Patients who Underwent Reoperation after Orthognathic Surgery : A 14-year Retrospective Study

자이스왈, 엠 샤리야 부산대학교 대학원 2023 국내석사

PURPOSE The purpose of this study was to analyze the patients clinically who underwent reoperationafter certain Orthognathic procedures, and to assess the reoperation rate. Furthermore, we also evaluated the incidence of intraoperative and postoperative complicationsthat led to the need for reoperation. METHODS Total 526 patients were selected who underwent Orthognathic surgery between July 2008 and February 2022 at the Department of Oral and Maxillofacial Surgery in Pusan National University Dental Hospital by a single surgeon. All the patients information was extracted from the electronic data base of our university. Demographic, radiologic, intraoperative and postoperative data were recorded and compiled. RESULTS Out of 526 patients, 265(50.3%) were males and 261(49.6%) were females. The total number of patients who showed complication is 89(16.9%) and the patients who underwent reoperation are 17(3.2%). The common complications that occurred were postoperative sensory disturbance (31; 5.8%), unwanted fractures (17; 3.2%), intraoperative nerve injury (11; 2%), wound dehiscence (11; 2%), infection (10; 1.9%), tooth injury (2; 0.3%) and others (18; 3.4%). The serious complications that led to reoperation include severe bleeding (6; 1.1%), unaesthetic results (5; 0.9%), non-union of maxilla (4; 0.7%) and, failed osteosynthesis (2; 0.3%). After 2018, all the orthognathic surgeries were performed with the help of Virtual Surgical Planning (VSP). After the application of VSP, the number of patients with complications statistically decreased. CONCLUSION The present study showed that the reoperation rate after orthognathic surgery was low, this rate was more decreased after applying 3D virtual surgery and 3D printed plate, especially in unaesthetic cases. 배경 및 목적 이번 연구의 목적은 악교정수술 후 재수술을 시행받은 환자들을 대상으로 임상적 분석을 시행하여, 재수술율과 재수술을 야기시키는 수술 중, 수술 후 합병증 유병율을 조사하고자 한다. 재료 및 방법 2008년부터 2022까지 부산대학교 구강악안면외과에서 한명의 술자에게 수술을 시행받은 526명의 환자가 조사되었다. 모든 환자의 정보는 전자차트에서 얻었으며, 환자의 신상정보, 방사선사진, 수술전후 데이터가 수집되었다. 결과 526명의 환자 중 265(50.3%)은 남자, 261(49.6%)는 여자였다. 합병증이 발생한 환자는 총 89명으로 16.9%의 환자에서 합병증이 발생하였다. 이중 17명(3.2%)이 재수술을 시행받았다. 가장 많이 발생한 합병증은 술후 감각이상(31, 5.8%), 원치않는 골분할(17,3.2%), 술중 신경손상(11.2%), 상처열개(11, 11.2%), 감염(10, 1.9%), 치아손상(2, 0.3%) 그리고 기타 합병증(18, 3.4%)이 있었다. 재수술이 필요한 심각한 합병증으로는 심한 출혈(6, 1.1%), 비심미적 결과(5, 0.9%), 상악골의 비유합(4, 0.7%), 그리고 하악골의 비유합(2, 0.3%) 으로 나타났다. 2018년 이후 모든 환자는 가상수술분석을 통해 수술이 계획되었고, 통계학적으로 합병증의 발생이 줄어들었다. 결론 이번 연구는 악교정 수술 후 재수술 비율은 3.2%로 높지 않았으며, 3차원 가상수술 적용 후 더욱 감소되었다.

(An) integrated system for simulation and image-guidance for orthognathic surgery

김대승 서울대학교 대학원 2013 국내박사

Accurate orthognathic surgery planning and planning transfer in the operation theatre are important factors for a successful surgery outcome with appropriate esthetical and functional improvements. However, conventional surgery has inherent limitations because the surgery planning is performed based on 2-dimensional cephalograms and the surgery operation depends on the experience and skill of surgeons. Moreover, 3-dimensional natural head position (NHP) information was not contained in not only the cephalogram but also CT data. To overcome these limitations, a new integrated system covering from virtual surgery planning to surgery guide system for orthognathic surgery has been developed. A new method for 3-dimensional simulation of NHP was also developed using POSIT algorithm. An accuracy evaluation was performed for the NHP simulation and the surgery guide system. In the virtual surgery planning system, realistic virtual osteotomy was performed on a 3-dimensional surface model reconstructed from the CT data using a haptic device. After designation of the landmarks on the surface model for a bone relocation procedure, bone segment was relocated by a combination of rotational and translational movements for each axis. The simulation of bone segment relocation was done according to the conventional surgery planning method. Movements of the designated landmarks were calculated for each relocation step and the final positions of the landmarks were compared with the conventional surgery plan to verify the simulation result. Relocated final position of the bone segment was recorded for the surgery guide system. For the NHP simulation, a photograph of subject was taken after attaching ceramic sphere markers on the subject’s maxillofacial area. CT data of the subject was also acquired while the ceramic markers were attached on the same positions. The POSIT algorithm was performed to acquire the relationship between the photograph coordinate system and the CT coordinate system. The 2-dimensional and 3-dimensional marker positions were acquired from the photograph and the CT data, respectively, and these positions were used for the POSIT algorithm calculation. The transformation matrix between the two coordinate systems was acquired as a result of the transformation calculation. Acquired transformation was applied to the CT data of the subject and the CT data was reoriented to the NHP. A Phantom was designed for an accuracy evaluation of the NHP simulation. Discrepancy between the simulated orientation and a virtual NHP of the phantom was determined as an accuracy of the simulation. The result of the accuracy evaluation of the NHP simulation showed clinically acceptable errors (degree) with 0.05±0.19 in roll, -0.56±0.37 in pitch and -0.01±0.29 in yaw. Intra-observer and inter-observer variation was evaluated for the NHP simulation and showed good agreement between measurements (p-value = 0.05). The surgery guide system tracked the maxilla bone segment intra-operatively using a 3-dimensional optical tracking system. Specially designed registration body was fabricated to improve the efficiency of the surgery. The point-to-point registration method was implemented to acquire the relationship between the optical tracking system coordinate system and the CT coordinate system. With the help of newly developed registration body, the registration procedure could be performed preoperatively. After the registration was completed, real-time maxilla bone segment position was tracked by the system. The result of the surgery planning which was previously acquired from the virtual surgery planning system was imported into the surgery guide system. During the tracking procedure, discrepancy between the planned position and the real-time intraoperative position of the maxilla bone segment was shown in 3-dimensional virtual environment. The discrepancy information of the points of interests on the bone segment was also quantitatively calculated. An accuracy of the surgery guide system was evaluated and compared with the conventional surgery guide method. The mean error was 0.47 ± 0.22 mm for the developed method and 1.05 ± 0.49 mm for the conventional method. The error of the developed method was significantly lower than the conventional method for all direction of movements. The accuracy of the surgery guide system was also evaluated in more clinical condition using the result of the model surgery. The CT data of the patient’s dental casts was acquired with the registration assembly. Registration between the CT coordinate system and the articulator coordinate system was performed, and the surface model of the dental cast was reoriented to the articulator coordinate system. The maxilla bone segment was relocated according to the model surgery result and tracked by the surgery guide system. Discrepancy of the landmark positions between the virtual surgery planned position and the actual position according to the model surgery was calculated and used for the evaluation of the system accuracy. Total of 7 cases of experiment was done and 5 measurements were implemented on 8 landmarks for each case. Total average of the absolute error ranged from 0.06 to 0.87 mm and the RMS error was 0.54±0.21 mm. The evaluation results showed clinically acceptable accuracy. In this study, integrated system for orthognathic surgery was developed. The system contained the virtual surgery planning system, the NHP simulation and the surgery guide system. The developed system helped the surgeons to predict the result of orthognathic surgery preoperatively and verify the exact intraoperative transfer of the surgery planning. In conclusion, this system was compatible with the clinical environment and increased the efficiency and the accuracy of the orthognathic surgery from the surgery planning to the surgery operation. The developed system is expected to be adopted in routine clinical situation and could be useful for the overall improvements of facial deformities after surgery.

Three-Dimensional Prediction of Facial Soft Tissue Changes after Orthognathic Surgery in Skeletal Class III Malocclusion Patients : A Deep Neural Network Approach Using CBCT Images

조현미 경희대학교 대학원 2025 국내박사

Introduction Predicting soft tissue changes in the maxillofacial region after orthognathic surgery still remains a challenge due to the complexity of soft tissue dynamics. This study aimed to develop a deep learning (DL) model using CBCT images of skeletal Class III malocclusion patients to predict three-dimensional (3D) changes in facial soft tissue following orthognathic surgery. Materials and Methods A total of 70 skeletal Class III malocclusion patients without significant asymmetry were included in this study. Using 3D Slicer, 3D facial soft tissue meshes were reconstructed from CBCT images obtained preoperatively and 1 year postoperatively. Soft tissue curvatures were simplified to generate 3D coordinate data, which, combined with the orthognathic surgery plan, served as input for the DL model. Of the total cohort, 64 patients were randomly assigned to the training group, where the DL model was designed and trained. The remaining 6 patients were allocated to the test group to assess the model’s prediction performance. Results A rendering engine was used to superimpose 3D facial meshes for visual validation. The comparison between preoperative and estimated meshes revealed similar changes, particularly in mandibular setback, despite slight differences in displacement, while postoperative and estimated meshes demonstrated high similarity. Analysis of 3D coordinate changes at facial soft tissue landmarks exhibited the highest similarity in vector changes at and Pog' and Me’, and lower similarity at Ala_L, Sn, and Sto. Vector distance analysis showed significant differences at specific landmarks at Sl. Overall, the DL model produced results closely matching actual outcomes, particularly in vertical facial height, nasal structure, and lateral aesthetic assessments based on anthropometric evaluation. Conclusion This study demonstrated the feasibility of incorporating a DL model with CBCT images in oral and maxillofacial surgery to predict three-dimensional facial soft tissue changes during orthognathic surgery, guided by the surgical plan. While the model achieved satisfactory overall accuracy, certain limitations were identified in regions characterized by detailed curvatures. Nevertheless, the model exhibited significant potential for visualizing surgical outcomes and supporting the development of patient-centered treatment. Continuous refinement of the model and the expansion of training data are anticipated to further enhance its performance and clinical applicability. 서론 악교정 수술 후의 안면 연조직 변화 예측은 구강악안면 연조직의 복합적인 특성으로 인해 여전히 도전적인 과제로 남아있다. 본 연구는 골격성 3급 부정교합 환자의 CBCT 영상을 활용하여 딥 러닝 기반 학습 모델을 개발하고, 이를 통해 악교정 수술 후 안면 연조직 변화를 3차원적으로 예측할 수 있는 가능성을 탐구하고자 하였다. 연구방법 현저한 비대칭이 관찰되지 않는 골격성 III급 부정교합 환자 70명을 대상으로, 3D Slicer를 사용하여 수술 전과 수술 후 1년의 CBCT 영상에서 안면 연조직의 3차원 메쉬를 형성하였다. 이후 연조직의 굴곡을 단순화하여 3차원 좌표 정보를 생성하고, 이를 악교정 수술계획과 함께 딥러닝 모델의 입력값으로 사용하였다. 무작위로 선택된 64명은 training group으로 분류되어, 이들 정보를 바탕으로 딥 러닝 모델을 설계하고 학습하였으며, 나머지 6명은 test group으로 분류하여 딥 러닝 모델의 예측 성능을 평가하였다. 결과 렌더링 엔진을 사용하여 3차원 메쉬를 중첩하여 시각적 검증을 수행하였다. 술 전과 예측 결과 비교에서 하악부의 후퇴 양상은 실제와 유사성을 보였으나, 이동량에서 소폭 차이가 관찰되었다. 술 후와 예측 결과를 비교하였을 때, 두 메쉬의 유사성을 확인하였다. 안면 연조직 계측점의 3차원 좌표 벡터 변화 분석에서 Pog’와 Me’에서 벡터 변화의 유사성이 가장 높았으며, Ala_L, Sn, 그리고 Sto에서는 유사성이 낮았다. 또한, 벡터의 크기 차이에서, Sl과 같은 특정 계측점에서는 차이가 있음을 확인하였다. 마지막으로, 계측점 기반의 수직 얼굴 길이, 코 구조 및 측면 심미성 평가에서 딥 러닝 모델은 전반적으로 실제와 유사한 결과를 나타내었다. 결론 본 연구는 악교정 수술 시 CBCT 영상과 딥러닝 모델을 활용하여 수술 계획에 따른 환자의 안면 연조직 변화를 3차원적으로 예측할 수 있는 가능성을 확인한 점에서 의의가 있다. 개발된 모델의 예측은 전반적으로 만족스러운 정확도를 나타냈으나, 정확한 하악의 후퇴량과 세밀한 변곡점의 예측에서는 다소 한계가 있음을 확인하였다. 본 연구의 결과는 수술 결과를 예측하고 시각화하여 환자 맞춤형 치료 계획을 제공할 수 있는 잠재력을 보여주었으며, 지속적으로 모델을 개선하고 학습 데이터를 확장함으로써 딥 러닝 모델의 성능 향상과 임상적 적용에 기여할 것으로 기대된다.

Correlation analysis of 3-dimensional changes of hard and soft tissues in Class III orthognathic surgery patients using CBCT

이태연 Graduate School, Yonsei University 2013 국내박사

Orthognathic surgery combined with orthodontics must enhance facial balance as well as skeletal harmony. Already a paradigm shift has occurred, with orthodontists and patients placing more importance on soft tissue esthetics. Because the facial soft tissues might not reflect the exact changes of the skeleton proportionally, an understanding of the soft tissue response due to skeletal changes from orthognathic surgery is essential during the treatment planning, prediction and treatment processes. The purposes of this study were to examine the correlation between changes of the soft and hard tissues 3-dimensionally in Class III orthognathic patients using CBCT and to provide indicators for the clinical diagnosis and treatment planning of orthognathic surgery. In this study, CBCT images were used within 2 weeks before surgery and at 6 months after surgery in adult Class III patients (9 males and 9 females, mean age 23.5 years) who underwent Le Fort I osteotomy and bilateral intraoral vertical ramus osteotomy (B-IVRO). Subjects were divided into Group 1 with no mandibular asymmetry (less than 2mm of menton deviation from the facial midline) and Group 2 with mandibular asymmetry (over than 4mm of menton deviation from the facial midline). CBCT images were reconstructed and 3D data of landmarks were calculated. Correlations and proportions of changes of the soft tissue to hard tissue were also obtained. 1. There were significant correlations of ΔB’ to ΔB and ΔPog’ to ΔPog in both groups on the horizontal axis (P<0.05). The proportions of ΔB’ to ΔB, and ΔPog’ to ΔPog in Group 1 were 0.53 and 0.67, respectively. Those of Group 2 were 0.77 and 0.88, respectively. 2. There were significant correlations of ΔB’ to ΔB, ΔPog’ to ΔPog and ΔMe’ to ΔMe on the sagittal axis in both groups (P<0.05). Correlation coefficients of ΔB’ to ΔB, ΔPog’ to ΔPog and ΔMe’ to ΔMe in Group 2 were greater than in Group 1. The proportions of ΔB’ to ΔB, ΔPog’ to ΔPog, and ΔMe’ to ΔMe in Group 1 were 0.94, 0.84, and 0.96, respectively. Those of Group 2 were 0.95, 0.91, and 1.03, respectively. 3. There were no significant correlations between changes in soft and hard tissues on the vertical axis in both groups (P>0.05). 4. There were significant correlations of the 3D distances between ΔB’ to ΔB, ΔPog’ to ΔPog and ΔMe’ to ΔMe in both groups (P<0.05). The proportions of ΔB’ to ΔB, ΔPog’ to ΔPog, and ΔMe’ to ΔMe in Group 1 were 0.92, 0.79, and 0.95, respectively. Those of Group 2 were 0.92, 0.88, and 0.97, respectively. In Class III orthognathic patients with mandibular asymmetry, the proportions of ΔB’ to ΔB, ΔPog’ to ΔPog, and ΔMe’ to ΔMe on the horizontal and sagittal axes and the 3D distances were different from patients with no mandibular asymmetry. Because sagittal changes of point A were little, proportions for changes of hard and soft tissues associated with the maxilla and upper lip were not calculated. Clinicians must consider these results in diagnosis and treatment planning of Class III mandibular asymmetry patients.