• Korean Journal of Head & Neck Oncology
• /
• v.36 no.1
• /
• pp.39-44
• /
• 2020

Xeroderma pigmentosum is a rare autosomal recessive disease, related to defects in DNA repair mechanism. It presents skin lesions on sun-exposed areas, leading to various skin cancer. Skin lesions can be treated with cryotherapy, skin resurfacing, 5-FU, Imiquimod, topical T4 endonuclease V, radiotherapy and genetic therapy, but invasive skin cancer should be treated by a surgery. We report a 12-year-old female xeroderma pigmentosum patient with recurrent basal cell carcinoma successfully treated by skin grafting. In that there is no cure for this disease, prevention and patient education is most important.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2019.10a
• /
• pp.931-934
• /
• 2019

Research using artificial intelligence to generate captions for an image has been studied extensively. However, these systems are unable to create creative stories that include more than one sentence based on image content. A story is a better way that humans use to foster social cooperation and develop social norms. This paper proposes a framework that can generate a relatively short story to describe based on the context of an image. The main contributions of this paper are (1) An unsupervised framework which uses recurrent neural network structure and encoder-decoder model to construct a short story for an image. (2) A huge English novel dataset, including horror and romantic themes that are manually collected and validated. By investigating the short stories, the proposed model proves that it can generate more creative contents compared to existing intelligent systems which can produce only one concise sentence. Therefore, the framework demonstrated in this work will trigger the research of a more robust AI story writer and encourages the application of the proposed model in helping story writer find a new idea.

• Journal of the Semiconductor & Display Technology
• /
• v.20 no.3
• /
• pp.6-10
• /
• 2021

Recognition of characters printed or marked on the PCB components from images captured using cameras is an important task in PCB components inspection systems. Previous optical character recognition (OCR) of PCB components typically consists of two stages: character segmentation and classification of each segmented character. However, character segmentation often fails due to corrupted characters, low image contrast, etc. Thus, OCR without character segmentation is desirable and increasingly used via deep neural networks. Typical implementation based on deep neural nets without character segmentation includes convolutional neural network followed by recurrent neural network (RNN). However, one disadvantage of this approach is slow execution due to RNN layers. LPRNet is a segmentation-free character recognition network with excellent accuracy proved in license plate recognition. LPRNet uses a wide convolution instead of RNN, thus enabling fast inference. In this paper, LPRNet was adapted for recognizing characters printed on PCB components with fast execution and high accuracy. Initial training with synthetic images followed by fine-tuning on real text images yielded accurate recognition. This net can be further optimized on Intel CPU using OpenVINO tool kit. The optimized version of the network can be run in real-time faster than even GPU.

• Phonetics and Speech Sciences
• /
• v.16 no.1
• /
• pp.25-31
• /
• 2024

This study examines the perceptual motivation behind dissimilation. Consistent with previous arguments suggesting that dissimilation originates from perception rather than production (Coetzee, 2005; Kiparsky, 2003; Scheer, 2013), we hypothesized that an oral stop with short of voice onset time (VOT) would be recognized as non-aspirated more often when it is followed by an aspirated stop with a long VOT. This hypothesis was tested through a perception experiment in which 32 Korean listeners made judgments on the first consonant of C₁VC₂V words manipulated with C₁ VOT and C₂ types. The results revealed that aspirated-based C₁ was recognized as aspirated or tense depending on the duration of VOT, while lenis-based C₁ was consistently recognized as lenis. The dissimilatory effect of aspirated C₂ was confirmed as anticipated, and furthermore, tense C₂ increased the ratio of tense responses more than aspirated C₂. These results provide evidence of a perceptual bias against recurrent aspirated stops, which may play a role in activating a dissimilatory rule or constraint in a language. The assimilatory effect of tense C₂ is in consistent with findings indicating that word-initial tensification is facilitated by the following tense stop in Korean (Kang & Oh, 2016; H. Kim, 2016).

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.15 no.7
• /
• pp.2304-2320
• /
• 2021

Intelligently detecting anomalies in health sensor data streams (e.g., Electrocardiogram, ECG) can improve the development of E-health industry. The physiological signals of patients are collected through sensors. Timely diagnosis and treatment save medical resources, promote physical health, and reduce complications. However, it is difficult to automatically classify the ECG data, as the features of ECGs are difficult to extract. And the volume of labeled ECG data is limited, which affects the classification performance. In this paper, we propose a Generative Adversarial Network (GAN)-based deep learning framework (called CAB) for heart arrhythmia classification. CAB focuses on improving the detection accuracy based on a small number of labeled samples. It is trained based on the class-imbalance ECG data. Augmenting ECG data by a GAN model eliminates the impact of data scarcity. After data augmentation, CAB classifies the ECG data by using a Bidirectional Long Short Term Memory Recurrent Neural Network (Bi-LSTM). Experiment results show a better performance of CAB compared with state-of-the-art methods. The overall classification accuracy of CAB is 99.71%. The F1-scores of classifying Normal beats (N), Supraventricular ectopic beats (S), Ventricular ectopic beats (V), Fusion beats (F) and Unclassifiable beats (Q) heartbeats are 99.86%, 97.66%, 99.05%, 98.57% and 99.88%, respectively. Unclassifiable beats (Q) heartbeats are 99.86%, 97.66%, 99.05%, 98.57% and 99.88%, respectively.

• Journal of Convergence for Information Technology
• /
• v.10 no.6
• /
• pp.156-163
• /
• 2020

This study made a final analysis of 184 people in their 40s and 50s living in South Gyeongsang Province to identify their relevance to oral health education experiences, oral health beliefs, oral efficacy, oral hygiene product use practices, and quality of life. When there was educational experience in oral efficacy and practice of oral hygiene product use according to oral health education, oral efficacy, oral hygiene product use practice and quality of life was high.In the relationship between oral health belief and oral efficacy, oral hygiene product use practice and quality of life, we could see that oral efficacy and quality of life were low when sensitivity and severity were high among oral health belief, and that oral efficacy and quality of life were high when importance, disability, and benefits were high. Middle age based on oral health care the results over the continuing and developers to develop customized education programs for different target for a longer term recurrent training and future.A Study on the Application through the program will be needed.

• Clinical and Experimental Pediatrics
• /
• v.64 no.10
• /
• pp.504-510
• /
• 2021

Population screening of newborns is an extremely important and informative diagnostic approach that allows early identification of babies who are predisposed to the development of a number of serious diseases. Some of these diseases are known and have effective treatment methods. Neonatal screening enables the early diagnosis and subsequent timely initiation of therapy. This helps to prevent serious complications and reduce the percentage of disability and deaths among newborns and young children. Primary immunodeficiency diseases and primary immunodeficiency syndrome (PIDS) are a heterogeneous group of diseases and conditions based on impaired immune system function associated with developmental defects and characterized by various combinations of recurrent infections, development of autoimmune and lymphoproliferative syndromes (genetic defects in apoptosis, gene mutation Fas receptor or ligand), granulomatous process, and malignant neoplasms. Most of these diseases manifest in infancy and lead to serious illness, disability, and high mortality rates. Until recently, it was impossible to identify children with PIDS before the onset of the first clinical symptoms, which are usually accompanied by complications in the form of severe coinfections of a viral-bacterial-fungal etiology. Modern advances in medical laboratory technology have allowed the identification of children with severe PIDS, manifested by T- and/or B-cell lymphopenia and other disorders of the immune system. This review discusses the main existing strategies and directions used in PIDS screening programs for newborns, including approaches to screening based on excision of T-cell receptors and kappa-recombination excision circles, as well as the potential role and place of next-generation sequencing technology to increase the diagnostic accuracy of these diseases.

• Korean Journal of Social Welfare Studies
• /
• v.48 no.4
• /
• pp.221-249
• /
• 2017

This study aims to investigate determinants that affect job retention of injured workers depending upon types of return to work in order to suggest define the intervention priority for those who returned to original works and for those who did not. After constructing explaining variables based on literature reviews, determinants were verified analyzing 1,292 people using Panel Study of Worker's Compensation Insurance(PSWCI) data. The job retention period turned out to be 46.6 months for those who returned to original work and 34.2 month for those who returned to new works. Injured workers who return to new works tend to have more unemployment experiences. As a result of Cox proportional regression analysis, the longer it takes to return to work, the longer both groups tend to retain after the accident. Age, recuperation period, health status, psycho-social rehabilitation, education and occupational training also affect on job retention probability for those who return to new work. Based upon the analyzed result, setting up an adequate duration for return-to-work, intervention for injured workers who experienced vulnerable working condition before the accident and continuous case management after return-to-work are suggested.

• 대한공업교육학회지
• /
• v.34 no.1
• /
• pp.19-43
• /
• 2009

An ideology what a modern society seeks is lifelong employment through lifelong learning. The single word which draws the attainment is lifelong vocational competency development. This study is designed for exploring its possibility and limit of a redesign of it, and tries to propose efficiently a redesign idea of a vocational educational path. To this end, it was analyzed by a subject for a national human resource foundation plan and a lifelong bocational competency development foundation plan. In addition to that, it was covered a structure and content analysis of main laws and regulations which are a base of policy making. An educational path could be divided into an area of formal education system, an area of workplace for labors, and an area of lifelong vocational competency development. And the insights were drawn by pulling apart an operational situation of main organizations and a recurrent process of the educational path. The redesign plan of a vocational education path we did analyze to activate lifelong vocational competency development system is deduced as following. Firstly, it should be re-organized with a direction towards an effective manpower training to be linked with skills and demands required by an industry. Secondly, it should be generalized for a model of recurrence between work and learning and its alternation for lifelong vocational competency development. Thirdly, it should create a cluster in the region for lifelong vocational competency development so that it is to form an efficient network to have a division of labor and its liaison with cooperative relationship.

• Asian Pacific Journal of Cancer Prevention
• /
• v.15 no.20
• /
• pp.8661-8666
• /
• 2014

Background: Docetaxel and cisplatin in combination with fluorouracil (DCF) regimen is accepted to be one of the standard regimens in the treatment of advanced gastric cancer. However, substantial toxicity has limited its use in daily clinical practice. Therefore, modification of DCF regimens, including introduction of capecitabine has been investigated to improve the safety profiles. In the present study, the efficacy and toxicity of a regimen with a modified dose of docetaxel and cisplatin in combination with oral capecitabine (DCX) was evaluated in untreated patients with HER2-negative advanced gastric cancer. Materials and Methods: Fifty-four patients with HER2-negative locally advanced or metastatic gastric cancer were included in this cohort. Patients received docetaxel $60mg/m^2$ plus cisplatin $60mg/m^2$ (day 1) combined with capecitabine $1650mg/m^2$ (days 1-14) every 3 weeks. Treatment response, survival, and toxicity were retrospectively analyzed. Results: The median age was 54 years (range: 24-76). The majority of patients (70%) had metastatic disease, while 11 patients (21%) had recurrent disease and underwent curative gastrectomy, and 5 patients (9%) had locally advanced disease (LAD). The median number of DCX cycles was 4. There were 28 partial responses and 11 complete responses, with an overall response rate of 72%. Curative surgery could be performed in four patients among five with LAD. At the median follow-up of 10 months, the median progression-free survival (PFS) and overall survival (OS) of the entire cohort of patients were 7.4 and 12.1 months, respectively. Dose modification was done in 12 patients due to toxicity in 8 and noncompliance in 4 patients. The most common hematological toxicity was neutropenia, which occurred at grade 3-4 intensity in 10 of 54 patients (27.7%). Febrile neutropenia was diagnosed only in two cases. Conclusions: DCX regimen offers prominent anti-tumor activity and considered to be effective first-line treatment with manageable toxicity for patients with HER2-negative advanced gastric cancer.