• The Korean Journal of Cytopathology
• /
• v.11 no.1
• /
• pp.1-10
• /
• 2000

The purpose of this study is to describe the cellular characteristics of endometrial hyperplasia without/with atypia in cervical smears. These cellular features were compared with those of normal endometrium and endometrial carcinoma. We reviewed 265 cervical smears : 64 normal proliferative endometrium, 118 endometrial hyperplasia without atypia, 21 endometrial hyperplasia with atypia, and 62 endometrial adenocarcinoma. Of these smears, 72(27.2%) smears which had diagnostic endometrial epithelial cells were selected for this study. The cytologic abnormalities about cellularity, background, changes in cellular architecture, alterations in nuclear size, anisokaryosis, chromatin pattern, nucleoli, cytoplasmic vacuoles, and mitosis were observed. Nuclear enlargement(1.6 to 2 times of the nucleus in the intermediate squamous cell) and anisokaryosis(${\geq}$2 fold in size variation) were highly suggestive of endometrial hyperplasia without/with atypia. The nuclei from endometrial hyperplasia with atypia were more coarsely granular in chromatin patterns than hyperplasia without atypia(33.3% vs 3.4%). Micronucleoli were observed in all endometrial conditions, but the presence of macronucleoli were more suggestive of hyperplasia with atypia(22.2%) and adenocarcinoma(55%). The changes in cellular architecture(loss of polarity, uneven internuclear distance, overlapping and loose arrangement) were seen in hyperplasia with atypia and adenocarcinoma. Characteristically, bloody background was seen in endometrial hyperpiasia, and cellular detritus or granular proteinaceous material was only observed in endometrial adenocarcinoma. Mitoses were also observed in adenocarcinoma. In conclusion, although there is no single parameter useful for the cytologic differential diagnosis of endometrial lesions, combined cytologic evaluation can be used to diagnose hyperplasia cytologically.

• 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.

• 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 the Korean Institute of Landscape Architecture
• /
• v.44 no.6
• /
• pp.13-25
• /
• 2016

This study aims to analyze the change process in order to set the improvement strategies for the zoo in the Seoul Children's Grand Park. The zoo can be reviewed through three significant time periods with noticeable changes. As a framework to analyze the major changes that happened in the zoo, this study looks at the changes in terms of the planning aspect as well as the animal welfare and program operation aspect. The findings are as follows: first, the era of general theme park turned out to have focused on exhibiting animals to meet visitor demands by expanding the zoo area of the zoo without enlarging stockyards for the animals. Second, the environmental park era created a zoo having entertaining and educational functions by arranging animal houses with the concept of zoological taxonomy and introducing animal behavioral enrichment, animal welfare programs and visitor participatory programs. The era of the zoo as an Urban Cultural Park improved old animal houses and facilities for the welfare of the animals and increased educational programs to preserve species and provide environmental education. The current status of the zoo turns out not to meet the conditions for creating an ecological zoo, which is the overall goal for contemporary zoos. The improvement strategies based on the analysis through three different eras are three-fold. First, the zoo needs to improve the boundary conditions of the animals to showcase animal wildness through landscape immersion. Second, the zoo should provide a shared environment for animals from the same habitats by changing the classification methods from the existing polyphyletic taxon to a classification that considers ecological habitat. Third, the zoo needs to develop various ecological education programs by supplementing specialists in professional education.

• Journal of Bio-Environment Control
• /
• v.32 no.1
• /
• pp.64-71
• /
• 2023

Crassulacean acid metabolism (CAM) plants use surplus CO₂ generated by cooling and heating at night when ventilation is not needed in a greenhouse. Schlumbergera truncata 'Pink Dew' is a multi-flowering cactus that needs more phylloclades for high-quality production. This study examined photosynthetic characteristics by the phylloclade levels of S. truncata in a growth chamber and a greenhouse for use of night CO₂ enrichment. The CO₂ uptake rate of the S. truncata's top phylloclade in a growth chamber exhibited a C₃ pattern, and the second phylloclade exhibited a C₃-CAM pattern. The CO₂ uptake rate of the top phylloclade in a greenhouse showed a negative value both day and night, but those of the second phylloclade exhibited a CAM pattern. The stomatal conductance and water-use efficiency (WUE) of S. truncata at both the top and second phylloclades were higher in a growth chamber than in a greenhouse. The WUE of S. truncata in a growth chamber and a greenhouse was higher at the second phylloclade, which is a CAM pattern compared with those of the top phylloclade. The daily total net CO₂ uptake of S. truncata was higher in a growth chamber than in a greenhouse. The daily total net CO₂ uptake of S. truncata at the second phylloclade had the highest value of 155 mmol·m^-2·d^-1 in a growth chamber. The night total CO₂ uptake of S. truncate at the second phylloclade was 3-fold higher in a growth chamber than in a greenhouse. S. truncata's second phylloclade exhibited a CAM pattern that uptake CO₂ at night, and the second phylloclade, was more mature than the top phylloclade. A multi-flowering cactus S. truncata 'Pink Dew' efficiently uptake night surplus CO₂ in the proper environmental condition with matured phylloclade.

• Korean Journal of Soil Science and Fertilizer
• /
• v.40 no.4
• /
• pp.274-279
• /
• 2007

The toxicitiy of organic sludge such as municipal sewage sludge (MSS), industrial sewage sludge (ISS), alcohol fermentation processing sludge (AFPS) and leather processing sludge (LPS) were evaluated with three environmental biomarkers as acetylcholinesterase, cytochrome P450, and heat shock protein 70 extracted from earthworm (Eisenia fetida). Their toxicities were compared with those of pig manure compost (PMC). MSS, ISS, LPS, and AFPS did not significantly affect the acetylcolinesterase activity, whereas only the elutriate of PMC slightly was increased the activity. MSS, AFPS, and PMC tended to slightly inhibit the cytochrome $P_{450}$ activity, but ISS and LPS showed significantly the inhibitory effect on cytochrome $P_{450}$. The hsp70 expression began to increase after treatments and showed high induction at 6 hour, followed by zero level at around 12 hour. The quantity of the hsp70 expressed by elutriate treatments of PMC, AFPS, MSS, ISS, and LPS was 1.9, 3.0, 3.3, 4.4, and 4.7 fold higher than that of distilled water. These results indicate that in toxicity tests of five organic waste materials, four kinds of sludge materials appeared more toxic than PMC. Results of AChE, P450, and hsp70 of earthworm might be useful for expecting or assessing an effect by exposure of organic wastes to earthworms in soil.

• Applied Biological Chemistry
• /
• v.40 no.5
• /
• pp.438-441
• /
• 1997

The relatedness of naturalized Bradyhizobium japonitum populations with soil physico-chemical characteristics as affected by paddy rice-upland soybean rotation cropping with conventional and none fertilization in Chilgog clay loam soils were determined as follows. The populations of B. japonicum in soils were increased from about $10^1$ in continuous paddy upto $10^1cells/g.soil$ only in one-year rotation of upland use with soybean cropping. Compared to the densities in plots of conventional fertilization, those in none fertilization were high ranging from 1.9 to 10 fold in 2-year upland use rotation and both in 3-year upland use rotation and 4-year upland use, respectively. The populations were positively correlated with soil organic matter $contents(r=0.83^*),\;Ca/K(r=0.74^*),\;and(Ca+Mg)/K(r=0.72^*)$ and were negatively correlated with soil $hardness(r=-0.73^*)$. And the soil populations increased by paddy-upland rotation resulted in superior symbiotic potentials to those in continuous paddy use in terms of nodule mass, nitrogenase activity, and soy-bean shoot dry weight.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.16 no.2
• /
• pp.927-939
• /
• 2015

It appears that the rapid advance in technology has allowed to broaden the variety of rail systems technology, thereby fostering new business opportunity in rail industry. The direction of rail systems operations is mainly two fold. In one direction, long distance operations between mega cities are pursued with help of high speed trains under development. In the other case, relatively short distance operations for covering intra-city or suburban area are becoming popular. A good example of the latter case is light rail transit (LRT) systems. Due to the short distance operation, it is thus expected that both the development and operation cost for LRT systems be reduced to some extent. The cost reduction desired in there can be gained by scaling down the sizes of both the trains and stations as compared to those of normal rail systems. However, it is not well known how the LRT stations can be scaled down. The objective of this paper is to study on how to slim down the stations (particularly, the functions room) of LRT systems. To achieve the objective, an approach is studied based on a modified method of design structure matrix (DSM). Specifically, using the enhanced DSM method, an integrated architecture is developed for the functions room, in which equipments are housed to perform the functions of electricity, signaling, and communication for LRT stations. The use of the result indicates that the desired reduction can be obtained with the approach taken in the paper.

• Proceedings of the Korean Vacuum Society Conference
• /
• 2013.08a
• /
• pp.91-91
• /
• 2013

Imagine a world where we could biomanufacture hybrid nanomaterials having atomic-scale resolution over functionality and architecture. Toward this vision, a fundamental challenge in materials science is how to design and synthesize protein-like material that can be fully self-assembled and exhibit information-specific process. In an ongoing effort to extend the fundamental understanding of protein structure to non-natural systems, we have designed a class of short peptides to fold like proteins and assemble into defined nanostructures. In this talk, I will talk about new strategies to drive the self-assembled structures designing sequence of peptide. I will also discuss about the specific interaction between proteins and inorganics that can be used for the development of new hybrid solar energy devices. Splitting water into hydrogen and oxygen is one of the promising pathways for solar to energy convertsion and storage system. The oxygen evolution reaction (OER) has been regarded as a major bottleneck in the overall water splitting process due to the slow transfer rate of four electrons and the high activation energy barrier for O-O bond formation. In nature, there is a water oxidation complex (WOC) in photosystem II (PSII) comprised of the earthabundant elements Mn and Ca. The WOC in photosystem II, in the form of a cubical CaMn4O5 cluster, efficiently catalyzes water oxidation under neutral conditions with extremely low overpotential (~160 mV) and a high TOF number. The cluster is stabilized by a surrounding redox-active peptide ligand, and undergo successive changes in oxidation state by PCET (proton-coupled electron transfer) reaction with the peptide ligand. It is fundamental challenge to achieve a level of structural complexity and functionality that rivals that seen in the cubane Mn4CaO5 cluster and surrounding peptide in nature. In this presentation, I will present a new strategy to mimic the natural photosystem. The approach is based on the atomically defined assembly based on the short redox-active peptide sequences. Additionally, I will show a newly identified manganese based compound that is very close to manganese clusters in photosystem II.

• Journal of Intelligence and Information Systems
• /
• v.27 no.3
• /
• pp.139-156
• /
• 2021

The biggest reason for using a deep learning model in image classification is that it is possible to consider the relationship between each region by extracting each region's features from the overall information of the image. However, the CNN model may not be suitable for emotional image data without the image's regional features. To solve the difficulty of classifying emotion images, many researchers each year propose a CNN-based architecture suitable for emotion images. Studies on the relationship between color and human emotion were also conducted, and results were derived that different emotions are induced according to color. In studies using deep learning, there have been studies that apply color information to image subtraction classification. The case where the image's color information is additionally used than the case where the classification model is trained with only the image improves the accuracy of classifying image emotions. This study proposes two ways to increase the accuracy by incorporating the result value after the model classifies an image's emotion. Both methods improve accuracy by modifying the result value based on statistics using the color of the picture. When performing the test by finding the two-color combinations most distributed for all training data, the two-color combinations most distributed for each test data image were found. The result values were corrected according to the color combination distribution. This method weights the result value obtained after the model classifies an image's emotion by creating an expression based on the log function and the exponential function. Emotion6, classified into six emotions, and Artphoto classified into eight categories were used for the image data. Densenet169, Mnasnet, Resnet101, Resnet152, and Vgg19 architectures were used for the CNN model, and the performance evaluation was compared before and after applying the two-stage learning to the CNN model. Inspired by color psychology, which deals with the relationship between colors and emotions, when creating a model that classifies an image's sentiment, we studied how to improve accuracy by modifying the result values based on color. Sixteen colors were used: red, orange, yellow, green, blue, indigo, purple, turquoise, pink, magenta, brown, gray, silver, gold, white, and black. It has meaning. Using Scikit-learn's Clustering, the seven colors that are primarily distributed in the image are checked. Then, the RGB coordinate values of the colors from the image are compared with the RGB coordinate values of the 16 colors presented in the above data. That is, it was converted to the closest color. Suppose three or more color combinations are selected. In that case, too many color combinations occur, resulting in a problem in which the distribution is scattered, so a situation fewer influences the result value. Therefore, to solve this problem, two-color combinations were found and weighted to the model. Before training, the most distributed color combinations were found for all training data images. The distribution of color combinations for each class was stored in a Python dictionary format to be used during testing. During the test, the two-color combinations that are most distributed for each test data image are found. After that, we checked how the color combinations were distributed in the training data and corrected the result. We devised several equations to weight the result value from the model based on the extracted color as described above. The data set was randomly divided by 80:20, and the model was verified using 20% of the data as a test set. After splitting the remaining 80% of the data into five divisions to perform 5-fold cross-validation, the model was trained five times using different verification datasets. Finally, the performance was checked using the test dataset that was previously separated. Adam was used as the activation function, and the learning rate was set to 0.01. The training was performed as much as 20 epochs, and if the validation loss value did not decrease during five epochs of learning, the experiment was stopped. Early tapping was set to load the model with the best validation loss value. The classification accuracy was better when the extracted information using color properties was used together than the case using only the CNN architecture.