• Journal of the Korea Convergence Society
• /
• v.12 no.5
• /
• pp.23-29
• /
• 2021

Recent deep learning-based natural language processing studies are conducting research to improve performance by training large amounts of data from various sources together. However, there is a possibility that the methodology of learning by combining data from various sources into one may prevent performance improvement. In the case of machine translation, data deviation occurs due to differences in translation(liberal, literal), style(colloquial, written, formal, etc.), domains, etc. Combining these corpora into one for learning can adversely affect performance. In this paper, we propose a new Corpus Weight Balance(CWB) method that considers the balance between parallel corpora in machine translation. As a result of the experiment, the model trained with balanced corpus showed better performance than the existing model. In addition, we propose an additional corpus construction process that enables coexistence with the human translation market, which can build high-quality parallel corpus even with a monolingual corpus.

• Journal of the Korea Convergence Society
• /
• v.12 no.11
• /
• pp.35-43
• /
• 2021

Recently, there has been a growing interest in zero-shot cross-lingual transfer, which leverages cross-lingual language models (CLLMs) to perform downstream tasks that are not trained in a specific language. In this paper, we point out the limitations of the data-centric aspect of quality estimation (QE), and perform zero-shot cross-lingual transfer even in environments where it is difficult to construct QE data. Few studies have dealt with zero-shots in QE, and after fine-tuning the English-German QE dataset, we perform zero-shot transfer leveraging CLLMs. We conduct comparative analysis between various CLLMs. We also perform zero-shot transfer on language pairs with different sized resources and analyze results based on the linguistic characteristics of each language. Experimental results showed the highest performance in multilingual BART and multillingual BERT, and we induced QE to be performed even when QE learning for a specific language pair was not performed at all.

• Journal of Korean Medicine Rehabilitation
• /
• v.33 no.4
• /
• pp.45-59
• /
• 2023

Objectives The aim of this study was to analyze the trends of researches on oriental medicine treatments for supraspinatus tendinopathy. Methods We used five databases for searching researches; Korean studies Information Service System, Oriental Medicine Advanced Searching Integrated System, Research Information Sharing Service, PubMed, and China National Knowledge Infrastructure. Only randomized controlled trials suitable for the subject were selected. The methodological quality of included randomized controlled trials (RCTs) was assessed by the Cochrane risk of bias tool. Results Twenty randomized controlled trials were analyzed. There were 9 types of treatment interventions; acupuncture, acupotomy, pharmacopuncture, electroacupuncture, fire needling, warming needle, catgut-embedding therapy, herbal medicine, cupping. The most frequently used treatment intervention was acupuncture and acupotomy. There were 9 outcome measurements including visual analog scale (VAS), Constant-Murley Score (CMS), and range of motion. The most used measurement was VAS. As a result of meta-analysis, acupuncture was more effective than control group in VAS. Additionally, acupotomy was clinically significant compared to control groups in VAS and CMS. Conclusions In this review, we analyzed researches on effectiveness of oriental medicine for supraspinatus tendinopathy. A provisional conclusion can be produced that acupuncture and acupotomy showed beneficial effect to supraspinatus tendinopathy. Although there were some RCT studies, many of them had a high risk of bias, so it is hard to conclude that our study can include overall clinical status. Further well-designed trials are needed.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.19 no.1
• /
• pp.137-142
• /
• 2024

Recently, deep learning technology has become those methods as de facto standards in the area of medical data representation. But, deep learning inherently requires a large amount of training data, which poses a challenge for its direct application in the medical field where acquiring large-scale data is not straightforward. Additionally, brain signal modalities also suffer from these problems owing to the high variability. Research has focused on designing deep neural network structures capable of effectively extracting spectro-spatio-temporal characteristics of brain signals, or employing self-supervised learning methods to pre-learn the neurophysiological features of brain signals. This paper analyzes methodologies used to handle small-scale data in emerging fields such as brain-computer interfaces and brain signal-based state prediction, presenting future directions for these technologies. At first, this paper examines deep neural network structures for representing brain signals, then analyzes self-supervised learning methodologies aimed at efficiently learning the characteristics of brain signals. Finally, the paper discusses key insights and future directions for deep learning-based brain signal analysis.

• Journal of The Korean Society of Clinical Toxicology
• /
• v.21 no.2
• /
• pp.135-142
• /
• 2023

Purpose: In patients with glufosinate poisoning, severe neurological symptoms may be closely related to a poor prognosis, but their appearance may be delayed. Therefore, this study aimed to determine whether the Acute Physiology and Chronic Health Evaluation II (APACHE II) score could predict the neurological prognosis in patients with glufosinate poisoning who present to the emergency room with alert mental status. Methods: This study was conducted retrospectively through a chart review for patients over 18 years who presented to a single emergency medical center from January 2018 to December 2022 due to glufosinate poisoning. Patients were divided into groups with a good neurological prognosis (Cerebral Performance Category [CPC] Scale 1 or 2) and a poor prognosis (CPC Scale 3, 4, or 5) to identify whether any variables showed significant differences between the two groups. Results: There were 66 patients (67.3%) with good neurological prognoses and 32 (32.8%) with poor prognoses. In the multivariate logistic analysis, the APACHE II score, serum amylase, and co-ingestion of alcohol showed significant results, with odds ratios of 1.387 (95% confidence interval [CI], 1.027-1.844), 1.017 (95% CI, 1.002-1.032), and 0.196 (95% CI, 0.040-0.948), respectively. With an APACHE II score cutoff of 6.5, the AUC was 0.826 (95% CI, 0.746-0.912). The cutoff of serum amylase was 75.5 U/L, with an AUC was 0.761 (95% CI, 0.652-0.844), and the AUC of no co-ingestion with alcohol was 0.629 (95% CI, 0.527-0.722). Conclusion: The APACHE II score could be a useful indicator for predicting the neurological prognosis of patients with glufosinate poisoning who have alert mental status.

• Journal of Embryo Transfer
• /
• v.22 no.3
• /
• pp.161-165
• /
• 2007

Oxytocin is a neurohypophyseal hormone which has multiple functions in mammals. Mainly, oxytocin regulates milk ejection and has an effect on uterine contraction and is related to maternal behavior. Maternal behavior is believed to be suppressed by stress and facilitated by oxytocin. In the cesarean section, oxytocin may be administrated into uterus to promote uterine involution. The present study aimed to test the effect of oxytocin into uterus on maternal behavior of rats with cesarean section. It was measured the effects on maternal behavior of oxytocin infused into uterus in rats with cesarean section as a stressor. In the first experiment, pup survival rate of between a control group and a group with laparotomy as a stress in natural parturition rats was compared. In the second experiment, survival rate for 2 weeks and maternal pup searching behavior (MPSV) were observed in one cesarean sectioned group without oxytocin and the other cesarean sectioned group with oxytocin. Infanticide was observed in stressed group in the first experiment while a normal maternal behavior was observed in a control one. In the second experiment, MPSV was only observed in a cesarean sectioned group with oxytocin and infanticide was observed in two groups except one rat which is thought to be affected by oxytocin as operated relatively late. This is the first study to show that the administration of oxytocin into uterus in the cesarean section is not involved in the regulation of maternal behavior in rats. In conclusion, this study proves the needs of oxytocin into brain in cesarean section related rats model and further study of maternal behavior list, like MPSV.

• Clinical and Experimental Pediatrics
• /
• v.50 no.9
• /
• pp.891-895
• /
• 2007

Purpose : Acute disseminated encephalomyelitis (ADEM) is a demyelinating disease of the central nervous system and mostly develops after viral illness or vaccinations. We investigated the clinical differences and neurologic outcomes according to the distribution of the lesions on brain MRI. Methods : The study group was composed of 21 patients from January 1995 to August 2003 in Kyunghee University hospital. We grouped the patients according to the MRI findings as follows. Group I (14 cases): Multi- or unifocal lesions only in the cerebral white matter. Group II (7 cases): lesions in the gray matter with or without white matter involvement. Results : 1. Preceding events were as follows: no defined prodrome (38.1%), upper respiratory tract infection (28.6%), nonspecific febrile illness (19.0%), gastointestinal disturbance and vaccination. 2. Presenting symptoms were as follows: seizures (76.2%), headache/vomiting (47.6%), altered consciousness (38.1%), hemiparesis, cerebellar ataxia, visual disturbance and facial nerve palsy. 3. Laboratory findings were as follows: CSF pleocytosis (76.2%), leucocytosis (38.1%) and elevated CSF protein (28.6%). 4. Fifteen patients were recovered completely without neurological sequelae. Three patients in group I and 1 patient in group II had intractable seizures. Two patients in group I and 2 patients in group II had motor disturbance. Conclusion : There were no statistically significant differences in preceding events, presenting symptoms, and neurological outcomes according to the distribution of the lesions on brain MRI. However, the ADEM have quite diverse clinical manifestations and neuroimage findings. MRI plays an important role in making diagnosis of the patients who are suspected of ADEM.

• Journal of Korean Society of Forest Science
• /
• v.108 no.3
• /
• pp.440-453
• /
• 2019

This study was carried out to investigate the effect of watching a two-dimensional (2D) forest video and a virtual reality (VR) forest video on stress reduction in adults. Experiments were conducted in an artificial climate room, and 40 subjects participated. After inducing stress in the subjects, subjects watched a 2D gray video, 2D forest video, or VR forest video for 5 mins. The autonomic nervous system activity was evaluated continuously in terms of measured heart rate variability during the experiment. After each experiment, the subject's psychological state was evaluated using a questionnaire. The 2D forest video decreased the viewer's stress index, increased HF, and reduced heart rate compared with the 2D gray video. The VR forest video had a greater stress index reduction effect, LF/HF increase effect, and heart rate reduction effect than the 2D gray video. Psychological measurements showed that subjects felt more comfortable, natural, and calm when watching the 2D gray video, 2D forest video or VR forest video. We also found that the 2D forest video and VR forest video increased positive emotions and reduced negative emotions compared to the 2D gray video. Based on these results, it can be concluded that watching the 2D forest and VR forest videos reduces the stress index and heart rate compared with watching the 2D gray video. Thus, it is considered that the 2D forest video increases the activity of the parasympathetic nervous system, and the VR forest video increases the activity of the sympathetic nervous system. The increased activity of the sympathetic nervous system upon watching the VR forest video is judged to be positive sympathetic nerve activity, such as novelty and curiosity, and not negative sympathetic activity, such as stress and tension. The results of this study are expected to be the basis for examining the visual effects of forest healing, with hope that the utilization of VR, the technology of the fourth industrial revolution in the forestry field, will broaden.

• Korean Journal of Food Science and Technology
• /
• v.51 no.2
• /
• pp.160-168
• /
• 2019

This study aimed to determine whether double-processed ginseng berry extract (PGBC) could improve learning and memory in an $A\hat{a}42$-induced Alzheimer's mouse model. Passive avoidance test (PAT) and Morris water-maze test (MWMT) were performed after mice were treated with PGBC, followed by acetylcholine (ACh) measurement and glial fibrillary acidic protein (GFAP) detection for brain damage. Furthermore, acetylcholinesterase (AChE) activity and choline acetyltransferase (ChAT) expression were analyzed using Ellman's and qPCR assays, respectively. Results demonstrated that PGBC contained a high amount of ginsenosides (Re, Rd, and Rg3), which are responsible for the clearance of $A{\hat{a}} 42$. They also helped to significantly improve PAT and MWMT performance in the $A{\hat{a}} 42-induced$ Alzheimer's mouse model when compared to the normal group. Interestingly, ACh and ChAT were remarkably upregulated and AChE activities were significantly inhibited, suggesting PGBC to be a palliative adjuvant for treating Alzheimer's disease. Altogether, PGBC was found to play a positive role in improving cognitive abilities. Thus, it could be a new alternative solution for alleviating Alzheimer's disease symptoms.

• Journal of Korea Water Resources Association
• /
• v.56 no.6
• /
• pp.403-417
• /
• 2023

In this research, a methodology was developed for constructing an appropriate rainfall image database for estimating rainfall intensity based on CCTV video. The database was constructed in the Large-Scale Climate Environment Chamber of the Korea Conformity Laboratories, which can control variables with high irregularity and variability in real environments. 1,728 scenarios were designed under five different experimental conditions. 36 scenarios and a total of 97,200 frames were selected. Rain streaks were extracted using the k-nearest neighbor algorithm by calculating the difference between each image and the background. To prevent overfitting, data with pixel values greater than set threshold, compared to the average pixel value for each image, were selected. The area with maximum pixel variability was determined by shifting with every 10 pixels and set as a representative area (180×180) for the original image. After re-transforming to 120×120 size as an input data for convolutional neural networks model, image augmentation was progressed under unified shooting conditions. 92% of the data showed within the 10% absolute range of PBIAS. It is clear that the final results in this study have the potential to enhance the accuracy and efficacy of existing real-world CCTV systems with transfer learning.