• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.22 no.4
• /
• pp.124-130
• /
• 2021

Dental prosthesis translucency importantly contributes to aesthetic outcomes. The purpose of this study was to investigate the effect of drying time and zirconia coloring agent type on translucency. For the experiment, 90 circular specimens were fabricated for each zirconia block. Then, zirconia specimens were treated with a coloring agent for 180 seconds and dried for 0 seconds (undried), 30 seconds (intermediate dry), or 30 minutes (complete dry). Then, a specimen was placed on a black standard tile or a white standard tile, and using a standard D65 light source reflected was measured using the light removal method. A total of three repeated measurements were obtained per specimen. One-way ANOVA was used to compare and analyze the relationship between zirconia translucency and drying time. Zirconia and coloring liquid types were significantly associated with translucency (P < 0.001). Although no significant difference was observed with respect to drying time (P > 0.922), zirconia in the completely dried (30 minutes) state was more translucent.

• Journal of the Korean Society of Propulsion Engineers
• /
• v.23 no.3
• /
• pp.119-126
• /
• 2019

For rockets using cryogenic liquid hydrogen or liquid oxygen, chilling is required to avoid cavitation and surge problems. Chilling is categorized by the initial chilling/filling stage and the low-temperature maintenance stage. In addition, to improve satellite insertion capability, a multi-ignition capability is required and accordingly chilling to prepare for the next ignition during low-gravity coasting is also required. This paper describes the overall aspects of filling and low temperature maintain marinating for the booster and the upper stage engine including chilling for multi-ignition.

• The Korean Journal of Applied Statistics
• /
• v.34 no.2
• /
• pp.153-166
• /
• 2021

A main goal of pharmacogenomics studies is to predict individual's drug responsiveness based on high dimensional genetic variables. Due to a large number of variables, feature selection is required in order to reduce the number of variables. The selected features are used to construct a predictive model using machine learning algorithms. In the present study, we applied several hybrid feature selection methods such as combinations of logistic regression, ReliefF, TurF, random forest, and LASSO to a next generation sequencing data set of 400 epilepsy patients. We then applied the selected features to machine learning methods including random forest, gradient boosting, and support vector machine as well as a stacking ensemble method. Our results showed that the stacking model with a hybrid feature selection of random forest and ReliefF performs better than with other combinations of approaches. Based on a 5-fold cross validation partition, the mean test accuracy value of the best model was 0.727 and the mean test AUC value of the best model was 0.761. It also appeared that the stacking models outperform than single machine learning predictive models when using the same selected features.

• Journal of Technologic Dentistry
• /
• v.43 no.2
• /
• pp.48-55
• /
• 2021

Purpose: This study aimed to analyze deformation according to post-curing of complete arch artificial teeth for temporary dentures printed with a digital light processing (DLP) printer. Methods: An edentulous model was prepared and an occlusal rim was produced. The edentulous model and occlusal rim were scanned using a model scanner. A complete denture was designed using a dental computer-aided design, and the denture base and artificial tooth were separated. Ten complete arch artificial teeth were printed using a 3D printer (DLP). Complete arch artificial teeth was classified into the following three groups: a group no post-curing (NC), a group with 10 minutes post-curing (10M), and a group with 20 minutes post-curing (20M). Specimens were scanned using a model scanner. The scanned data were overlapped with the reference data. Statistical analysis was performed using one-way ANOVA analysis of variance, Kruskal-Wallis test, and Mann-Whitney U test (α=0.05). Results: Regarding the overall deviation of complete arch artificial teeth, the NC group showed the lowest mean deviation of 111.13 ㎛ and the 20M group showed the highest mean deviation of 131.03 ㎛. There were statistically significant differences among the three groups (p<0.05). Conclusion: The complete arch artificial tooth showed deformation due to post-curing. In addition, the largest shrinkage deformation was observed at 10 minutes of post-curing, whereas the least deformation was observed at 20 minutes.

• Journal of Technologic Dentistry
• /
• v.42 no.4
• /
• pp.313-320
• /
• 2020

Purpose: The purpose of this study is to compare and analyze the accuracy of single crowns based on the type of occlusal surface. Methods: A single crown wax pattern was fabricated in three types of occlusal surface. The prepared wax pattern was replicated with silicone, and stone was injected to create a stone model. The prepared specimens were scanned using a model scanner. Scans were classified into three groups, and each scan was performed six times to analyze the trueness and precision of a single crown. In addition, only the occlusal surface area was analyzed for trueness and precision. Data were analyzed using the Kruskal-Wallis H test, a nonparametric test (α=0.05). Results: With regard to the trueness value of the occlusal scan area, the no occlusal tooth attrition (NA) group showed the largest error of 3.5 ㎛, and the complete occlusal tooth attrition (CA) group showed the lowest value of 3.1 ㎛. The NA group had the greatest precision, and the medium occlusal tooth attrition (MA) group and CA group showed a low precision value of 3.2 ㎛; the difference between the groups was statistically significant (α=0.05). In the color difference map, the CA group showed a lower error than the NA group. Conclusion: The occlusal surface with severe attrition had excellent accuracy, but the accuracy of the group without attrition was low. There were significant differences between groups, but clinically acceptable values were shown.

• Korean Journal of Remote Sensing
• /
• v.39 no.5_3
• /
• pp.949-965
• /
• 2023

In this study, we propose a method to monitor the surface area of agricultural reservoirs in South Korea using Sentinel-1 synthetic aperture radar images and the deep learning model, Swin Transformer. Utilizing the Google Earth Engine platform, datasets from 2017 to 2021 were constructed for seven agricultural reservoirs, categorized into 700 K-ton, 900 K-ton, and 1.5 M-ton capacities. For four of the reservoirs, a total of 1,283 images were used for model training through shuffling and 5-fold cross-validation techniques. Upon evaluation, the Swin Transformer Large model, configured with a window size of 12, demonstrated superior semantic segmentation performance, showing an average accuracy of 99.54% and a mean intersection over union (mIoU) of 95.15% for all folds. When the best-performing model was applied to the datasets of the remaining three reservoirsfor validation, it achieved an accuracy of over 99% and mIoU of over 94% for all reservoirs. These results indicate that the Swin Transformer model can effectively monitor the surface area of agricultural reservoirs in South Korea.

• Journal of Biomedical Engineering Research
• /
• v.23 no.5
• /
• pp.351-364
• /
• 2002

Conventional 3D ultrasound imaging using mechanical ID arrays suffers from poor elevation resolution due to the limited depth-of-focus (DOF). On the other hand, 3D imaging systems using 2D phased arrays have a large number of active channels and hence require a very expensive and bulky beamforming hardware. To overcome these limitations, a new real-time volumetric imaging method using curved 2-D arrays is presented, in which a small subaperture, consisting of 256 elements, moves across the array surface to scan a volume of interest. For this purpose, a 2-D curved array is designed which consists of 90$\times$46 elements with 1.5λ inter-element spacing and has the same view angles along both the lateral and elevation directions as those of a commercial mechanical 1-D array. In the proposed method, transmit and receive subapertures are constructed by cutting the four corners of a rectangular aperture to obtain a required image qualify with a small number of active channels. In addition the receive subaperture size is increased by using a sparse array scheme that uses every other elements in both directions. To suppress the grating lobes elevated due to the increase in clement spacing, fold-over array scheme is adopted in transmit, which doubles the effective size of a transmit aperture in each direction. Computer simulation results show that the proposed method can provide almost the same and greatly improved resolutions in the lateral and elevation directions, respectively compared with the conventional 3D imaging with a mechanical 1-D array.

• Journal of Sasang Constitutional Medicine
• /
• v.10 no.1
• /
• pp.143-160
• /
• 1998

PURPOSE This studied the biomechanical body segment parameters of Korean adults with reference to Sasang Constitutional Medicine(SCM). We anlyzed the characteristics of Sasang constitution through Body Measurement, Immersion Method, and Reaction Board Method. SUBJECT Subjects were 72 persons. There were made 49 and Female 23. Mean age was . And There were Taeumin 30, Soyangin 17, and Soeumin 25. METHOD The items of Body Measurement were 51. Limbs were measured right side. Volume was gained with Immersion Method and Weight was calculated with the density equation for limbs from Drills and Contini. Center of Mass was gained with Reaction Board Method. RESULT 1. In Body Measurement there were the significant differences with each constitution of man's circumference. 2. In Volume there were the significant differences between Taeumin and Soeumin in mand. And the volume rate of head and neck, hand, and foot was the lowest in Taeumin, but the highest in Soeumin. 3. In center of mass Soyangin was higher than others.

• Korean Journal of Remote Sensing
• /
• v.37 no.5_1
• /
• pp.1149-1161
• /
• 2021

With a trend of the utilization of computer vision for satellite images, cloud detection using deep learning also attracts attention recently. In this study, we conducted a U-Net cloud detection modeling using SPARCS (Spatial Procedures for Automated Removal of Cloud and Shadow) Cloud Dataset with the image data augmentation and carried out 10-fold cross-validation for an objective assessment of the model. Asthe result of the blind test for 1800 datasets with 512 by 512 pixels, relatively high performance with the accuracy of 0.821, the precision of 0.847, the recall of 0.821, the F1-score of 0.831, and the IoU (Intersection over Union) of 0.723. Although 14.5% of actual cloud shadows were misclassified as land, and 19.7% of actual clouds were misidentified as land, this can be overcome by increasing the quality and quantity of label datasets. Moreover, a state-of-the-art DeepLab V3+ model and the NAS (Neural Architecture Search) optimization technique can help the cloud detection for CAS500 (Compact Advanced Satellite 500) in South Korea.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2022.05a
• /
• pp.399-401
• /
• 2022

A definitive diagnosis to classify the breast malignancy status may be achieved by microscopic analysis using surgical open biopsy. However, this procedure requires experts in the specializing of histopathological image analysis directing to time-consuming and high cost. To overcome these issues, deep learning is considered practically efficient to categorize breast cancer into benign and malignant from histopathological images in order to assist pathologists. This study presents a pre-trained convolutional neural network model architecture with a 100% fine-tuning scheme and Adagrad optimizer to classify the breast cancer histopathological images into benign and malignant using a 40× magnification BreaKHis dataset. The pre-trained architecture was constructed using the InceptionResNetV2 model to generate a modified InceptionResNetV2 by substituting the last layer with dense and dropout layers. The results by demonstrating training loss of 0.25%, training accuracy of 99.96%, validation loss of 3.10%, validation accuracy of 99.41%, test loss of 8.46%, and test accuracy of 98.75% indicated that the modified InceptionResNetV2 model is reliable to predict the breast malignancy type from histopathological images. Future works are necessary to focus on k-fold cross-validation, optimizer, model, hyperparameter optimization, and classification on 100×, 200×, and 400× magnification.