SLSang-Hwy Lee

Sang-Hwy Lee

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


Sang-Hwy Lee has a total of 81 co-authors. Here are the top co-authors by number of shared publications:

  • Sang-Hoon Kang

    Department of Oral and Maxillofacial Surgery, National Health Insurance Service Ilsan Hospital, Goyang, Republic of Korea

    12 shared publications
  • Hak-Jin Kim

    Department of Oral and Maxillofacial Surgery, College of Dentistry, Yonsei University, Seoul, South Korea

    11 shared publications
  • Bong Chul Kim

    Department of Oral and Maxillofacial Surgery, Daejeon Dental Hospital, Wonkwang University College of Dentistry, Daejeon, Republic of Korea

    8 shared publications
  • Hye-Jin Tak

    Oral Science Research Center, College of Dentistry, Yonsei University, Seoul, South Korea

    6 shared publications
  • Ji Wook Choi

    Department of Oral and Maxillofacial Surgery, Oral Science Research Center, Yonsei University, College of Dentistry, Seoul, Republic of Korea

    6 shared publications
  • Jin Keun Seo

    Department of Computational Science and Engineering, Yonsei University, KOREA, 50 Yonsei-Ro, Seodaemun-Gu, Seoul, 120-749

    4 shared publications

Latest Publications

  • Open Access

    Mandibular ramus remodeling after vertical osteotomy: a three-dimensional quantitative and qualitative assessment with sequential segmental-marrow superimposition

    by Sang-Hwy Lee, Bong Chul Kim, Hoon Cho, Nam-Kyoo Kim, Jong-Ki Huh, Hyung Seog Yu, Sang-Hoon Kang

    The intraoral vertical ramus osteotomy (IVRO) is a surgical technique for prognathic or asymmetric mandible. We wanted to quantitatively and qualitatively analyze the three-dimensional (3D) remodeling pattern of the mandibular ramus one year after IVRO. We reconstructed 3D mandibular models using 3D CT data from 108 IVRO hemi-mandibles of 54 subjects. The immediate postoperative (Imm) proximal and distal segments were separated and individually superimposed to one-year postoperative (Post) mandible by employing a direct segmental-isolated marrow superimposition. Analysis showed extensive regional bone resorption and apposition resulting in a remodeled Post ramus; the more overlapping patterns of group of A (lateral type) and B (lateral/proximal type) suffered resorption-dominant remodeling compared with the less-overlapped group C (proximal type) and D (gap type), while the level of apposition was similar for all groups. Post ramus showed different levels of sigmoid notch filling with newly-formed bones and various shapes of new bony projections at the posterior border of the medial ramus. Bone resorption in group A was significantly and negatively correlated with protrusion delta (the difference between the postoperative and preoperative protrusive mandibular movement ranges), and bone apposition was positive with amount of postoperative maximum opening. All results indicate dynamic remodeling of the Imm ramus after IVRO in relation to the segmental overlapping pattern to form a smooth and continuous skeletal surface. The new superimposition method of direct segmental overlapping could detect the detailed remodeling pattern.


  • Open Access

    Automatic 3D Registration of Dental CBCT and Face Scan Data Using 2D Projection Images

    by Hyoung Suk Park, Chang Min Hyun, Sang-Hwy Lee, Jin Keun Seo, Kiwan Jeon

    This paper presents a fully automatic registration method of dental cone-beam computed tomography (CBCT) and face scan data. Difficulties in accurately merging facial scans and CBCT images arise from the different image acquisition methods and limited area of correspondence between the two facial surfaces. In addition, it is difficult to use machine learning techniques because they use face-related 3D medical data with radiation exposure, which are difficult to obtain for training. The proposed method addresses these problems by reusing an existing machine-learning-based 2D landmark detection algorithm in an open-source library and developing a novel mathematical algorithm that identifies paired 3D landmarks from knowledge of the corresponding 2D landmarks. A main contribution of this study is that the proposed method does not require annotated training data of facial landmarks because it uses a pre-trained facial landmark detection algorithm that is known to be robust and generalized to various 2D face image models. Note that this reduces a 3D landmark detection problem to a 2D problem of identifying the corresponding landmarks on two 2D projection images generated from two different projection angles. Here, the 3D landmarks for registration were selected from the sub-surfaces with the least geometric change under the CBCT and face scan environments. For the final fine-tuning of the registration, the Iterative Closest Point method was applied, which utilizes geometrical information around the 3D landmarks. The proposed method achieved an averaged mean surface distance error of 0.74 mm for three pairs of CBCT and face scan datasets.


  • Open Access

    Three-Dimensional Change of Lip after Two-Jaw Surgery in Facial Asymmetry Using Facial Scanner

    by Young-Jae Kim, Sung-Hwan Choi, Yoon Jeong Choi, Kee-Joon Lee, Sang-Hwy Lee, Hyung-Seog Yu

    A facial scanner and three-dimensional computed tomography (CT) were used to evaluate the three-dimensional change in lip asymmetry before and after two-jaw surgery for 22 patients with facial asymmetry (menton deviation > 3 mm). We used the labrale superius (Ls), deviated/non-deviated-side cheilions (Ch-D/Ch-ND), and labrale inferius (Li) to construct the upper and lower lip planes to evaluate the lip asymmetry. A correlation analysis was performed to determine the factors related to the vertical change in the cheilions (ΔChZ-D/ND). In the transverse axis, Ch-D and Li moved to improve the asymmetry after surgery. All landmarks, except the Ls, moved backward in the anteroposterior axis. In the vertical axis, significant upward movement was observed in all hard tissue landmarks; however, there were no significant changes in the soft tissue. In the lip plane, the difference in the height of Ch-D and Ch-ND was significantly reduced (1.38 mm vs. 0.72 mm). In the anteroposterior axis, the ΔChZ-D/ND showed significant correlations with the mandibular setback. In the vertical axis, the ΔCh-ND showed significant correlations with the maxillary impaction of the non-deviated side. The improvement in lip asymmetry post-surgery was mainly achieved by the movement of the lower lip and Ch-D rather than the upper lip and Ch-ND.

  • Open Access

    Reduction malarplasty using a simulated surgical guide for asymmetric/prominent zygoma

    by Sang-Hoon Kang, Hye-Jin Tak, Hak-Jin Kim, Sang-Hwy Lee

    The present study introduces a reduction malarplasty using a three-dimensional (3D)-printed surgical guide and evaluates the guide’s technical applicability. Twenty malarplasties were performed for 12 subjects with zygomatic asymmetry/prominency using the current method. 3D reconstruction of the craniomaxillofacial region and fine dental occlusion was made with image data from computed tomograpy and dental scanning. A computer-assisted surgical simulation was performed for reduction malarplasty and a surgical guide was designed for later 3D printing. The manufactured surgical guide was introduced to the operation field to guide the surgery; its surgical accuracy was confirmed by comparing five corresponding points from preoperative simulation and postoperative data. We successfully performed the reduction malarplasty with the surgical guide. The accuracy level of surgery fell to 0.93 mm of total median difference for the corresponding zygoma points of preoperative simulations and postoperative zygoma. The anterior and upper points showed less error level (0.59 and 0.73 mm difference, respectively) than did other points. We developed a computer-assisted surgical technique using a surgical guide for asymmetrical/prominent zygoma which proved to be simple, practical, and accurate; it is expected to help surgeons perform reduction malarplasty with ease and accuracy.


  • Open Access

    3D cephalometric landmark detection by multiple stage deep reinforcement learning

    by Sung Ho Kang, Kiwan Jeon, Sang-Hoon Kang, Sang-Hwy Lee

    The lengthy time needed for manual landmarking has delayed the widespread adoption of three-dimensional (3D) cephalometry. We here propose an automatic 3D cephalometric annotation system based on multi-stage deep reinforcement learning (DRL) and volume-rendered imaging. This system considers geometrical characteristics of landmarks and simulates the sequential decision process underlying human professional landmarking patterns. It consists mainly of constructing an appropriate two-dimensional cutaway or 3D model view, then implementing single-stage DRL with gradient-based boundary estimation or multi-stage DRL to dictate the 3D coordinates of target landmarks. This system clearly shows sufficient detection accuracy and stability for direct clinical applications, with a low level of detection error and low inter-individual variation (1.96 ± 0.78 mm). Our system, moreover, requires no additional steps of segmentation and 3D mesh-object construction for landmark detection. We believe these system features will enable fast-track cephalometric analysis and planning and expect it to achieve greater accuracy as larger CT datasets become available for training and testing.


  • Open Access

    Fully-customized distraction assembly for maxillofacial distraction osteogenesis: a novel device and its experimental accuracy verification

    by Sang-Hoon Kang, Hye-Jin Tak, Ha-Won Park, Jin-Ung Kim, Sang-Hwy Lee

    Background: A new distraction osteogenesis assembly system comprising a fully customized CAD/CAM-based fixation unit and ready-made distraction unit was developed. The aim of this study was to introduce our new distraction system and to evaluate its accuracy level in a sampled mandibular distraction osteogenesis.Methods: Our system consists of a fully customized CAD/CAM-based fixation plate unit with two plates for each moving and anchoring part, and a ready-made distraction unit with attachment slots for fixation plates. The experimental distractions were performed on 3D-printed mandibles for one control and two experimental groups (N = 10 for each group). All groups had reference bars on the chin region and teeth to measure distraction accuracy. The control group had the classical ready-made distraction system, and experimental groups 1 and 2 were fitted with our new distraction assembly using a different distractor-positioning guide design. All distracted experimental mandibles were scanned by CT imaging, then superimposed on a 3D simulation to get their discrepancy levels.Results: The measured 3D distances between the reference landmarks of the surgical simulations and the experimental surgeries for the three groups were significantly different (p < 0.0001) by statistical analysis. The errors were greater in the control group (with a total average of 19.18 ± 3.73 mm in 3D distance between the simulated and actual reference points) than those in the two experimental groups (with an average of 3.68 ± 1.41 mm for group 1 and 3.07 ± 1.39 mm for group 2). The customized distraction assembly with 3D-printed bone plate units in group 1 and 2, however, did not show any significant differences between simulated and actual distances (p > 0.999).Conclusion: Our newly-developed distraction assembly system with CAD/CAM plate for the distraction osteogenesis of the mandible produced a greater level of accuracy than that of a conventional distraction device. The system appears to address existing shortcomings of conventional distraction devices, including inaccuracy in vector-controlled movement of the system. However, it also needs to be further developed to address the requirements and anatomical characteristics of specific regions.


  • Open Access

    Learning-based local-to-global landmark annotation for automatic 3D cephalometry

    by Hye Sun Yun, Tae Jun Jang, Sung Min Lee, Sang-Hwy Lee, Jin Keun Seo

    The annotation of three-dimensional (3D) cephalometric landmarks in 3D computerized tomography (CT) has become an essential part of cephalometric analysis, which is used for diagnosis, surgical planning, and treatment evaluation. The automation of 3D landmarking with high-precision remains challenging due to the limited availability of training data and the high computational burden. This paper addresses these challenges by proposing a hierarchical deep-learning method consisting of four stages: 1) a basic landmark annotator for 3D skull pose normalization, 2) a deep-learning-based coarse-to-fine landmark annotator on the midsagittal plane, 3) a low-dimensional representation of the total number of landmarks using variational autoencoder (VAE), and 4) a local-to-global landmark annotator. The implementation of the VAE allows two-dimensional-image-based 3D morphological feature learning and similarity/dissimilarity representation learning of the concatenated vectors of cephalometric landmarks. The proposed method achieves an average 3D point-to-point error of 3.63 mm for 93 cephalometric landmarks using a small number of training CT datasets. Notably, the VAE captures variations of craniofacial structural characteristics.