• The Sea:JOURNAL OF THE KOREAN SOCIETY OF OCEANOGRAPHY
• /
• v.29 no.1
• /
• pp.1-13
• /
• 2024

The southwestern part of the East Sea is known to have a high primary productivity compared to those in the northern and eastern parts, which is attributed to nutrients supplies either by Tsushima Warm Current or by coastal upwelling. However, research on the biological pump in this area is limited. We developed machine learning models to estimate net community production (NCP), a measure of biological pump, with high spatial and time scales of 4 km and 8 days, respectively. The models were fed with the input parameters of sea surface temperature, chlorophyll-a, mixed layer depths, and photosynthetically active radiation and trained with observed NCP derived from high resolution measurements of surface O₂/Ar. The root mean square error between the predicted values by the best performing machine model and the observed NCP was 6 mmol O₂ m^-2 d^-1, corresponding to 15% of the average of observed NCP. The NCP in the central part of the Ulleung Basin was highest in March at 49 mmol O₂ m^-2 d^-1 and lowest in June and July at 18 mmol O₂ m^-2 d^-1. These seasonal variations were similar to the vertical nitrate flux based on the ³He gas exchange rate and to the particulate organic carbon flux estimated by the ²³⁴Th disequilibrium method. To expand this method, which produces NCP estimate for spring and summer, to autumn and winter, it is necessary to devise a way to correct bias in NCP by the entrainment of subsurface waters during the seasons.

• Korean Journal of Remote Sensing
• /
• v.40 no.2
• /
• pp.191-202
• /
• 2024

Global warming is causing abnormal climates worldwide due to the increase in greenhouse gas concentrations in the atmosphere, negatively affecting ecosystems and humanity. In response, various countries are attempting to reduce greenhouse gas emissions in numerous ways, and interest in blue carbon, carbon absorbed by coastal ecosystems, is increasing. Known to absorb carbon up to 50 times faster than green carbon, blue carbon plays a vital role in responding to climate change. Particularly, the tidal flats of South Korea, one of the world's five largest tidal flats, are valued for their rich biodiversity and exceptional carbon absorption capabilities. While previous studies on blue carbon have focused on the carbon storage and annual carbon absorption rates of tidal flats, there is a lack of research linking tidal flat area changes detected using satellite data to carbon storage. This study applied the direct difference water index to high-resolution satellite data from PlanetScope and RapidEye to analyze the area and changes of the Nakdong River estuary tidal flats over six periods between 2013 and 2023, estimating the carbon storage for each period. The analysis showed that excluding the period in 2013 with a different tidal condition, the tidal flat area changed by up to approximately 5.4% annually, ranging from about 9.38 km² (in 2022) to about 9.89 km² (in 2021), with carbon storage estimated between approximately 30,230.0 Mg C and 31,893.7 Mg C.

• Journal of the Korean Society of Radiology
• /
• v.81 no.1
• /
• pp.21-40
• /
• 2020

Magnetic resonance imaging (MRI) is an essential modality for the diagnosis of musculoskeletal system defects because of its higher soft-tissue contrast and spatial resolution. With the recent development of MRI-related technology, faster imaging and various image plane reconstructions are possible, enabling better assessment of three-dimensional musculoskeletal anatomy and lesions. Furthermore, the image quality, diagnostic accuracy, and acquisition time depend on the MRI protocol used. Moreover, the protocol affects the efficiency of the MRI scanner. Therefore, it is important for a radiologist to optimize the MRI protocol. In this review, we will provide guidance on patient positioning; selection of the radiofrequency coil, pulse sequences, and imaging planes; and control of MRI parameters to help optimize the MRI protocol for the six major joints of the musculoskeletal system.

• Korean Journal of Radiology
• /
• v.25 no.1
• /
• pp.55-61
• /
• 2024

Objective: This study aimed to evaluate the safety and efficacy of intranodal lymphangiography and thoracic duct embolization (TDE) for chyle leakage (CL) after thyroid surgery. Materials and Methods: Fourteen patients who underwent intranodal lymphangiography and TDE for CL after thyroid surgery were included in this retrospective study. Among the 14 patients, 13 underwent bilateral total thyroidectomy with neck dissection (central compartment neck dissection [CCND], n = 13; left modified radical neck dissection (MRND), n = 11; bilateral MRND, n = 2), and one patient underwent left hemithyroidectomy with CCND. Ten patients (76.9%) had high-output CL (> 500 mL/d). Before the procedure, surgical intervention was attempted in three patients (thoracic duct ligation, n = 1; lymphatic leakage site ligation, n = 2). Lymphangiographic findings, technical and clinical successes, and complications were analyzed. Technical success was defined as the successful embolization of the thoracic duct after access to the lymphatic duct via the transabdominal route. Clinical success was defined as the resolution of CL or surgical drain removal. Results: On lymphangiography, ethiodized oil leakage near the surgical bed was identified in 12 of 14 patients (85.7%). The technical success rate of TDE was 78.6% (11/14). Transabdominal antegrade access was not feasible due to the inability to visualize the identifiable cisterna chyli or a prominent lumbar lymphatic duct. Among patients who underwent a technically successful TDE, the clinical success rate was 90.1% (10/11). The median time from the procedure to drain removal was 3 days (with a range of 1-13 days) for the 13 patients who underwent surgical drainage. No CL recurrence was observed during the follow-up period (ranging from 2-44 months; median, 8 months). There were no complications, except for one case of chylothorax that developed after TDE. Conclusion: TDE appears to be a safe and effective minimally invasive treatment option for CL after thyroid surgery, with acceptable technical and clinical success rates.

• Journal of the Korean Regional Science Association
• /
• v.40 no.2
• /
• pp.107-130
• /
• 2024

Urban land surface temperature (LST) change is a major environmental factor that affects the thermal comfort, energy consumption, and health of urban residents. Most studies that explored the relationship between LST and urban built-environment form analyzed only midday LST. This study explores the diurnal variation of summertime LST in Seoul using ECOSTRESS data, which observes LST at various times of the day and analyzes whether the LST variation differs by built environment type. Launched in 2018, ECOSTRESS operates in a non-sun-synchronous orbit, observing LST with a high resolution of 70 meters. This study collected data from early morning (6:25) to evening (17:26) from 2019 to 2022 to build time-series LST. Based on greenery, water bodies, and building form data, eight types of Seoul's built environment were derived by hierarchical clustering, and the LST fluctuation characteristics of each cluster were compared. The results showed that the spatial disparity in LST increased after dawn, peaked at noon, and decreased again, highlighting areas with rapid versus stable LST changes. Low-rise and high-rise compact districts experienced fast, high temperature increases and high variability, while low-density apartments experienced moderate LST increases and low variability. These results suggest urban forms that can mitigate rapid daytime heating.

• The Journal of the Acoustical Society of Korea
• /
• v.43 no.2
• /
• pp.243-252
• /
• 2024

Among the Foley sound generation models that have recently begun to be studied, a sound generation technique using the Vector Quantized-Variational AutoEncoder (VQ-VAE) structure and generation model such as Pixelsnail are one of the important research subjects. On the other hand, in the field of deep learning-based acoustic signal compression, residual vector quantization technology is reported to be more suitable than the conventional VQ-VAE structure. Therefore, in this paper, we aim to study whether residual vector quantization technology can be effectively applied to the Foley sound generation. In order to tackle the problem, this paper applies the residual vector quantization technique to the conventional VQ-VAE-based Foley sound generation model, and in particular, derives a model that is compatible with the existing models such as Pixelsnail and does not increase computational resource consumption. In order to evaluate the model, an experiment was conducted using DCASE2023 Task7 data. The results show that the proposed model enhances about 0.3 of the Fréchet audio distance. Unfortunately, the performance enhancement was limited, which is believed to be due to the decrease in the resolution of time-frequency domains in order to do not increase consumption of the computational resources.

• Journal of the Korean Association of Geographic Information Studies
• /
• v.27 no.1
• /
• pp.115-127
• /
• 2024

This research aimed to construct models with various structures based on the Transformer module and to perform land cover classification, thereby examining the applicability of the Transformer module. For the classification of land cover, the Unet model, which has a CNN structure, was selected as the base model, and a total of four deep learning models were constructed by combining both the encoder and decoder parts with the Transformer module. During the training process of the deep learning models, the training was repeated 10 times under the same conditions to evaluate the generalization performance. The evaluation of the classification accuracy of the deep learning models showed that the Model D, which utilized the Transformer module in both the encoder and decoder structures, achieved the highest overall accuracy with an average of approximately 89.4% and a Kappa coefficient average of about 73.2%. In terms of training time, models based on CNN were the most efficient. however, the use of Transformer-based models resulted in an average improvement of 0.5% in classification accuracy based on the Kappa coefficient. It is considered necessary to refine the model by considering various variables such as adjusting hyperparameters and image patch sizes during the integration process with CNN models. A common issue identified in all models during the land cover classification process was the difficulty in detecting small-scale objects. To improve this misclassification phenomenon, it is deemed necessary to explore the use of high-resolution input data and integrate multidimensional data that includes terrain and texture information.

• The Korean Journal of Nuclear Medicine Technology
• /
• v.13 no.3
• /
• pp.56-60
• /
• 2009

Purpose: The various studies and efforts to develop program are in progress in the field of nuclear medicine for the purpose of reducing scan time. The Oncoflash is one of the programs used in whole body bone scan which allows to maintain the image quality while to reduce scan time. When Those applications are used in clinical setting, both the image quality and reduction of scan time should be considered, therefore, the purpose of this study was to determine the criteria for proper scan speed. Materials and Methods: The subjects of this study were the patients who underwent whole body bone scan at the departments of nuclear medicine in the Asan Medical Center located in Seoul from 1st to 10th, July, 2008. The whole body bone images obtained in the scan speed of 30cm/min were classified by the total counts into under 800 K, and over 800 K, 900 K, 1,000 K, 1,500 K, and 2,000 K. The image quality were assessed qualitatively and the percentages of those of 1,000K and under of total counts were calculated. The FWHM before and after applying the Oncoflash were analyzed using images obtained in $^{99m}Tc$ Flood and 4-Quadrant bar phantom in order to compare the resolution according to the amount of total counts by the application of the Oncoflash. Considering the counts of the whole body bone scan, the dosed 2~5 mCi were used. 152 patients underwent the measurement in which the counts of Patient Postioning Monitor (PPM) were measured with including head and the parts of chest which the starting point of whole body bone scan from 7th to 26th, August, 2008. The correlations with total counts obtained in the scan speed of 30cm/min among them were analyzed (The exclusion criteria were after over six hours of applying isotopes or low amount of doses). Results: The percentage of the whole body bone image which has the geometric average of total counts of under 1,000K among them obtained in the scan speed of 30cm/min were 17.6%(n=58) of 329 patients. The qualitative analysis of the image groups according to the whole body counts showed that the images of under 1,000K were assessed to have coarse particles and increased noises. The analysis on the FWHM of the images before and after applying the Oncoflash showed that, in the case of PPM counts of under 3.6 K, FWHM values after applying the Oncoflash were higher than that before applying the Oncoflash, whereas, in the case of that of over 3.6 K, the FWHM after applying the Oncoflash were not higher than that before applying the Oncoflash. The average of total counts at 2.5~3.0 K, 3.1~3.5 K, 3.6~4.0 k, 4.1~4.5 K, 4.6~5.0 K, 5.1~6.0 K, 6.1~7.0 K, and 7.1 K over (in PPM) were $965{\pm}173\;K$, $1084{\pm}154\;K$, $1242{\pm}186\;K$, $1359{\pm}170\;K$, $1405{\pm}184\;K$, $1640{\pm}376\;K$, $1,771{\pm}324\;K$, and $1,972{\pm}385\;K$, respectively and the correlations between the counts in PPM and the total counts of image obtained in the scan speed of 30 cm/min demonstrated strong correlation (r=.775, p<.01). Conclusions: In the case of PPM coefficient over 3.6 K, the image quality obtained in the scan speed of 30cm/min and after applying the Oncoflash was similar to that obtained in the scan speed of 15 cm/min. In the case of total counts over 1,000 K, it is expected to reduce scan time without any damage on the image quality. In the case of total counts under 1,000 K, however, the image quality were decreased even though the Oncoflash is applied, so it is recommended to perform the re-image in the scan speed of 15 cm/min.

• The Korean Journal of Nuclear Medicine Technology
• /
• v.16 no.2
• /
• pp.18-24
• /
• 2012

Purpose : Although lung ventilation SPECT (LV-SPECT) has a good sensitivity in detection of deep lung lesions, it is difficult to apply the LV-SPECT to patients having breathing problems due to limited examination time. In this study, we evaluated the usefulness of LEAP collimator, which provides high detection sensitivity and tolerable resolution, for the LV-SPECT in terms of diagnostic accuracy and examination time. Materials and Methods : Four volunteers inhaled Technegas (370 MBq) and the lung ventilation planar scan (LVPS, 300 counts/view (cpv)) with LEHR collimator was performed using Siemens E.cam scanner as a reference test. LV-SPECT scans were performed with three collimators, LEHR, LEUHR, and LEAP, in low (7 kcpv) and high (70 kcpv) counting modes. The count ratios of left (LT) and right (RT) lung segments were calculated on the geometric mean view of anterior and posterior images for LVPS and on the summed coronal images of LV-SPECT, respectively. Comparing to LVPS, the usefulness of three different collimators for LV-SPECT was evaluated through statistical analysis (paired t-test), on count ratios of lung segments. Results : The average LT:RT ratio in LVPS was 47:53. For LV-SPECT, there were negligible difference of the LT:RT ratios (48:52 on average) among three different collimators in low and high counting modes. Comparing to standard LVPS with LEHR, all LV-SPECTs with different collimators resulted in similar diagnostic accuracy through paired t-test (p>0.05). The scan time in LVPS (6 views) was 17.3 min. For LV-SPECT (128 views) in low counting mode, it took 18.7 (LEUHR), 15.0 (LEHR), and 12.3 min (LEAP), respectively. Conclusion : Comparing to standard LVPS, the LV-SPECT with LEAP in low counting mode provided the comparable diagnostic accuracy in addition to shortened scan time.

• Korean Journal of Agricultural and Forest Meteorology
• /
• v.8 no.1
• /
• pp.1-9
• /
• 2006

Regardless of the recent observed warmer winters in Korea, more freeze injuries and associated economic losses are reported in fruit industry than ever before. Existing freeze-frost forecasting systems employ only daily minimum temperature for judging the potential damage on dormant flowering buds but cannot accommodate potential biological responses such as short-term acclimation of plants to severe weather episodes as well as annual variation in climate. We introduce 'dormancy depth', in addition to daily minimum temperature, as a complementary criterion for judging the potential damage of freezing temperatures on dormant flowering buds of grape vines. Dormancy depth can be estimated by a phonology model driven by daily maximum and minimum temperature and is expected to make a reasonable proxy for physiological tolerance of buds to low temperature. Dormancy depth at a selected site was estimated for a climatological normal year by this model, and we found a close similarity in time course change pattern between the estimated dormancy depth and the known cold tolerance of fruit trees. Inter-annual and spatial variation in dormancy depth were identified by this method, showing the feasibility of using dormancy depth as a proxy indicator for tolerance to low temperature during the winter season. The model was applied to 10 vineyards which were recently damaged by a cold spell, and a temperature-dormancy depth-freeze injury relationship was formulated into an exponential-saturation model which can be used for judging freeze risk under a given set of temperature and dormancy depth. Based on this model and the expected lowest temperature with a 10-year recurrence interval, a freeze risk probability map was produced for Hwaseong County, Korea. The results seemed to explain why the vineyards in the warmer part of Hwaseong County have been hit by more freeBe damage than those in the cooler part of the county. A dormancy depth-minimum temperature dual engine freeze warning system was designed for vineyards in major production counties in Korea by combining the site-specific dormancy depth and minimum temperature forecasts with the freeze risk model. In this system, daily accumulation of thermal time since last fall leads to the dormancy state (depth) for today. The regional minimum temperature forecast for tomorrow by the Korea Meteorological Administration is converted to the site specific forecast at a 30m resolution. These data are input to the freeze risk model and the percent damage probability is calculated for each grid cell and mapped for the entire county. Similar approaches may be used to develop freeze warning systems for other deciduous fruit trees.