3D 모델 기반 기하처리를 이용한 치아 임플란트 식립 계획 수립 및 유한요소분석

박형욱(Hyungwook Park),박철우(Chulwoo Park),박상진(Sangjin Park),김명수(Myung-Soo Kim),박형준(Hyungjun Park) (사)한국CDE학회 2012 한국 CAD/CAM 학회 학술발표회 논문집 Vol.2012 No.2

치아 임플란트 시술의 성공률을 높이기 위해서는 시술 전 적절한 식립 계획을 수립하는 것이 매우 중요하다. 최근에는 의료영상 및 3D 데이터 기반 S/W가 시술계획 수립 과정에 널리 활용되고 있으나, 식립 위치와 방향 선정이 시술자 경험과 주관적 판단에 의존되어 상당한 편차가 존재한다. 본 논문에서는 식립 위치 근처에 존재하는 주변치아 및 구강에 따라 3D 모델의 기하학적 처리를 통해 임플란트 식립 위치 및 방향을 결정하고, 이를 유한요소분석을 통해 입증함으로써 보다 일관된 식립 계획을 수립하는 방안을 제시한다. 제안된 방안에서는 시술부위의 주변 치아들에 대해 PCA 알고리즘을 적용하여 얻어진 최소포함원기등(Minimum Enclosing Cylinder)을 이용하여 식립위치 및 방향을 결정한다. 시술부위 주변 원기둥의 윗면 중심점과 원기둥 반경을 고려하여 식립 위치를 결정하고, 주변 원기둥의 중심축에 가중치를 고려한 평균을 통해 식립 방향을 결정한다. 또한 시술부위 주변 원기둥들과 신경관 모델 간의 간섭을 고려하여 시술부위 원기둥의 최대 반경과 최대 길이를 추정한 후, 이를 임플란트 크기 및 유형 선정에 활용한다. 유한요소분석을 통해 식립 계획 정보의 적합성을 확인한 다음, 식립계획 정보를 토대로 식립보조도구(surgical stents)의 3D 모델을 생성한다. 최종적으로 3D 치아 임플란트 시술 시뮬레이션을 수행함으로써 제시된 방안이 치아 임플란트 시술용 교육컨텐츠에 유용하게 활용됨은 물론 식립계획 수립을 위한 유용한 정보를 제공할 수 있음을 보인다.

<i>In-Situ</i> Synchrotron X-Ray Scattering Study of Thin Film Growth by Atomic Layer Deposition

Park, Yong Jun,Lee, Dong Ryeol,Lee, Hyun Hwi,Lee, Han-Bo-Ram,Kim, Hyungjun,Park, Gye-Choon,Rhee, Shi-Woo,Baik, Sunggi American Scientific Publishers 2011 Journal of Nanoscience and Nanotechnology Vol.11 No.2

<P>We report an atomic layer deposition chamber for in-situ synchrotron X-ray scattering study of thin film growth. The chamber was designed for combined synchrotron X-ray reflectivity and two-dimensional grazing-incidence X-ray diffraction measurement to do a in-situ monitoring of ALD growth. We demonstrate ruthenium thermal ALD growth for the performance of the chamber. 10, 20, 30, 50, 70, 100, 150 and 250-cycled states are measured by X-ray scattering methods during ALD growth process. Growth rate is calculated from thickness values and the surface roughness of each state is estimated by X-ray reflectivity analysis. The crystal structure of initial growth state is observed by Grazing-incidence X-ray diffraction. These results indicate that in-situ X-ray scattering method is a promising analysis technique to investigate the initial physical morphology of ALD films.</P>

A new approach for optimal single biarc fitting

Hyungjun Park(박형준) (사)한국CDE학회 2004 한국 CAD/CAM 학회 학술발표회 논문집 Vol.2004 No.2

Tangible Augmented Prototyping of Digital Handheld Products

Hyungjun Park,Hee-Cheol Moon (사)한국CDE학회 2007 한국CAD/CAM학회 국제학술발표 논문집 Vol.2007 No.8

Semi-supervised Deep Learning Extraction for Information of Lung Cancer Staging from Unstructured Reports of PET-CT Interpretation

( Hyungjun Park ),( Chang-min Choi ) 대한결핵 및 호흡기학회 2021 대한결핵 및 호흡기학회 추계학술대회 초록집 Vol.129 No.-

Purpose Information about lung cancer stage is crucial but requires laborious annotation. We developed a deep learning model for extracting information from PET-CT reports for determining lung cancer stages. Methods PET-CT reports were acquired from two cohorts of patients with lung cancer who were diagnosed at a tertiary hospital between January 2004 and March 2020. One cohort of 20,466 PET-CT reports was used for training and the validation set, and the other cohort of 4190 PET-CT reports was used for the extra-validation set. A preprocessing model (Lung Cancer Spell Checker) was applied to correct the typos, and pseudo-labeling was used for training the model. The deep learning model was constructed using the Convolutional-Recurrent Neural Network. The performance metrics for the prediction model were accuracy, precision, sensitivity, micro-AUROC, and AUPRC. Results For the extraction of primary lung cancer location, the model showed a micro-AUROC of 0.913 and 0.946 in the validation set and the extra-validation set, respectively. For metastatic lymph nodes, the model showed a sensitivity of 0.827 and a specificity of 0.960. In predicting distant metastasis, the model showed a micro-AUROC of 0.944 and 0.950 in the validation and the extravalidation set, respectively. (Figure 1) Conclusion Our deep learning model could extract lung cancer stage information from PET-CT reports and may facilitate lung cancer studies by alleviating laborious annotation by clinicians.

Note on Tangible AR Interaction based on Fingertip Touch using Small-sized Markers

Hyungjun Park,Ho-Kyun Jung,Sang-Jin Park (사)한국CDE학회 2013 한국CAD/CAM학회 국제학술발표 논문집 Vol.2010 No.8

Although big-sized markers are good for accurate marker recognition and tracking, they are easily occluded by other objects and deteriorate natural visualization and level of immersion during user interaction in AR environments. In this paper, we propose an approach to exploiting the use of rectangular markers to support tangible AR interaction based on fingertip touch using small-sized markers. It basically adjusts the length, width, and interior area of rectangular markers to make them more suitably fit to longish objects like fingers. It also utilizes convex polygons to resolve the partial occlusion of a marker and properly enlarges the pattern area of a marker while adjusting its size without deteriorating the quality of marker detection. We obtained encouraging results from users that the approach is accurate and tangible enough to support a pseudo feeling of touching virtual products with human hands or fingertips during design evaluation of digital handheld products.

Adaptive B-spline volume representation of measured BRDF data for photorealistic rendering

Park, Hyungjun,Lee, Joo-Haeng Society for Computational Design and Engineering 2015 Journal of computational design and engineering Vol.2 No.1

Measured bidirectional reflectance distribution function (BRDF) data have been used to represent complex interaction between lights and surface materials for photorealistic rendering. However, their massive size makes it hard to adopt them in practical rendering applications. In this paper, we propose an adaptive method for B-spline volume representation of measured BRDF data. It basically performs approximate B-spline volume lofting, which decomposes the problem into three sub-problems of multiple B-spline curve fitting along u-, v-, and w-parametric directions. Especially, it makes the efficient use of knots in the multiple B-spline curve fitting and thereby accomplishes adaptive knot placement along each parametric direction of a resulting B-spline volume. The proposed method is quite useful to realize efficient data reduction while smoothing out the noises and keeping the overall features of BRDF data well. By applying the B-spline volume models of real materials for rendering, we show that the B-spline volume models are effective in preserving the features of material appearance and are suitable for representing BRDF data.

Simultaneous improvement of the dielectric constant and leakage currents of ZrO₂ dielectrics by incorporating a highly valent Ta⁵⁺ element

Park, Bo-Eun,Oh, Il-Kwon,Park, Jong Seo,Seo, Seunggi,Thompson, David,Kim, Hyungjun The Royal Society of Chemistry 2018 Journal of Materials Chemistry C Vol.6 No.36

<P>Dynamic random access memory (DRAM) is reaching the scaling limit owing to the requirements for a high capacitance density and low leakage current of metal-insulator-metal (MIM) capacitors. We investigated the Ta-doped ZrO2 dielectric as a novel high-<I>k</I> candidate, utilizing the precise control of Ta-doping concentration using the atomic layer deposition (ALD) supercycle process. We systematically studied the chemical composition and crystal structure of ALD Ta-doped ZrO2 and its effects on the electrical properties of MIM capacitors. It was shown that ZrO2 becomes more stoichiometric with the introduction of Ta, which is attributed to the suppression of oxygen vacancy (VO) formation. The change in the atomic arrangement due to the substitution of Zr with Ta and the reduction of VO enhances the crystallinity of the cubic phase and causes a decrease in the molar volume of the ZrO2 films. As a result of the change in the crystal structure along with the high dielectric polarizability of Ta, the dielectric constant of ALD Ta-doped ZrO2 increases by up to 80% compared to that of undoped ZrO2 films. Moreover, the reduction of the VO species suppresses the emission of the carriers, which lowers the leakage current density by two orders of magnitude in the Ta-doped ZrO2 films as compared to that in the undoped ZrO2 films. In general, the dielectric constant and leakage currents have a trade-off relationship in a single high-<I>k</I> dielectric system. Consequently, proper doping of Ta into ZrO2 using ALD is a promising solution to overcome the technical limits of conventional high-<I>k</I> dielectrics.</P>

Visualization of the Etiology of Pleural Effusion Using Contrastive-loss

( Hyungjun Park ),( Chang-min Choi ) 대한결핵 및 호흡기학회 2021 대한결핵 및 호흡기학회 추계학술대회 초록집 Vol.129 No.-

Background Blood and fluid analysis is widely used for classifying the etiology of pleural effusion. However, most studies have been focused on simply determining the presence of the disease. The aim of this research was to classify the pleural effusion etiology with deep learning models using contrastive loss. Methods All patients with pleural effusion who underwent thoracentesis between January 2009 and December 2019 at Asan Medical Center were retrospectively analyzed. Five different models (multinomial logistic regression, random forest, gradient boost, deep neural network, contrastive-loss) were compared for categorizing the etiology of pleural effusion. The performance metrics were top-1 accuracy, top-2 accuracy, and micro- and weighted-AUROC. To visualize the embedding space of the contrastive-loss model, UMAP and t-SNE were used. Results While the five models showed similar performance in the validation set, the contrastive-loss model had the highest accuracy in the extra-validation set. Specifically, the accuracy and micro-AUROC of the contrastive-loss model were 81.7% and 0.942 in the validation set, and 66.2% and 0.867 in the extra-validation set. (Table 1) The embedding space visualization in the contrastive-loss model revealed the typical and atypical effusion Results by comparing true and false positives of rule-based criteria. (Figure 1) Conclusion The contrastive-loss model could be used for classifying the etiology of pleural effusion. With visualization of the contrastiveloss model, clinicians may obtain useful insights for diagnosing the etiology by distinguishing between typical and atypical disease types.

Knowledge-Based CNC Torch Path Generation for Laser Cutting of Planar Shapes

Hyungjun Park,Dong-Gyu Ahn (사)한국CDE학회 2005 한국CAD/CAM학회 국제학술발표 논문집 Vol.2005 No.8