• Annals of Hepato-Biliary-Pancreatic Surgery
• /
• v.27 no.3
• /
• pp.251-257
• /
• 2023

Backgrounds/Aims: This is a retrospective analysis of whether the 8th edition American Joint Committee on Cancer (AJCC) was a significant improvement over the 7th AJCC distal extrahepatic cholangiocarcinoma classification. Methods: In total, 111 patients who underwent curative resection of mid-distal bile duct cancer from 2002 to 2019 were included. Cases were re-classified into 7th and 8th AJCC as well as clinicopathological univariate and multivariate, and Kaplan-Meier survival curve and log rank were calculated using R software. Results: In patient characteristics, pancreaticoduodenectomy/pylorus preserving pancreaticoduodenectomy had better survival than segmental resection. Only lymphovascular invasion was found to be significant (hazard ratio 2.01, p = 0.039) among all clinicopathological variables. The 8th edition AJCC Kaplan Meier survival curve showed an inability to properly segregate stage I and IIA, while there was a large difference in survival probability between IIA and IIB. Conclusions: The 8th distal AJCC classification did resolve the anatomical issue with the T stage, as T1 and T3 showed improvement over the 7th AJCC, and the N stage division of the N1 and N2 category was found to be justified, with poorer survival in N2 than N1. Meanwhile, in TMN staging, the 8th AJCC was able differentiate between early stage (I and IIA) and late stage (IIB and III) to better explain the patient prognosis.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.24 no.2
• /
• pp.197-203
• /
• 2024

Food allergies are common and accidents occur annually. However, many people lack knowledge of the severity of allergies and food ingredients. Allergy-related applications currently on the market have problems such as providing information by relying only on certain certified products, food ingredients, and barcodes. This design plans a customized service application for food allergy patients. In this application, after extracting the text of the image using OCR technology, the food ingredients were read and displayed in large letters. In addition, if the user selects an ingredient that cannot be consumed through filtering technology, the restricted food is quickly and conveniently shown when searching for food ingredients. Finally, when scanning a barcode or searching for a product, food ingredient information is provided through barcode scanning and search engine technology that provides ingredient information of the product. Therefore, the purpose of this paper is to design an app in which users with food allergies can easily check food ingredients and avoid allergic reactions using databases and various information search methods.

• The Korean Journal of Franchise Management
• /
• v.15 no.2
• /
• pp.57-75
• /
• 2024

Purpose: Nowadays, many restaurants use serving robots. Initially, many people thought that Covid-19 caused the spread of serving robots. However, even as the endemic, many restaurants still use serving robots. Therefore, this study examines why many customers choose restaurants with serving robots, using the UTAUT2 framework. Additionally, this study explores whether shyness has a moderating effect on these factors. Research design, data and methodology: Data were collected from 307 consumers who had visited a restaurant using a serving robot and analyzed using SmartPLS 4.0 software. A total of 286 datasets were analyzed. Result: We found that the precedence factors of UTAUT2 (Performance Expectancy, Effort Expectancy, Social Influence, Facilitating Condition, Hedonic Motivation) had a positive effect on attitude. Furthermore, attitude had a significant positive effect on Behavioral Intention. However, shyness did not appear to have a moderating effect among these factors. This is likely due to customers using serving robots for very short time, as identified in the literature review. Conclusions: As a result of this study, it was explained that Hedonic Motivation had the most significant positive effect on shaping attitudes toward restaurants using serving robots through the UTAUT2 model.

• Journal of Advanced Technology Convergence
• /
• v.3 no.1
• /
• pp.43-49
• /
• 2024

The beauty industry has evolved into a highly competitive service sector, expanding alongside economic growth. Nail art, in particular, has witnessed a remarkable surge in popularity. To efficiently manage individual time, nail salons have adopted reservation systems. However, the issue of no-shows, where customers fail to keep their appointments, has emerged as a significant societal and economic concern. This study seeks to identify and address the root causes of no-shows in nail salons, providing essential data for the industry. A comprehensive survey was conducted, involving 476 customers from nail salons in the Daejeon and Chungcheong regions. Survey responses were analyzed using SPSS 24.0 software. Results revealed that a change of mind was the primary reason for no-shows. This underscores the importance of promoting reservation awareness among both businesses and consumers. To mitigate no-shows, suggestions included implementing non-refundable reservation deposits and imposing cancellation fees. Notably, the beauty industry currently lacks systematic regulations concerning no-shows, highlighting the necessity for future research aimed at developing reservation-related guidelines and standards.

• Journal of Preventive Medicine and Public Health
• /
• v.57 no.3
• /
• pp.298-303
• /
• 2024

Objectives: The aim of this study was to estimate drug prescription indicators in outpatient services provided at Iran Social Security Organization (SSO) healthcare facilities. Methods: Data on all prescribed drugs for outpatient visits from 2017 to 2018 were extracted from the SSO database. The data were categorized into 4 main subgroups: patient characteristics, provider characteristics, service characteristics, and type of healthcare facility. Logistic regression models were used to detect risk factors for inappropriate drug prescriptions. SPSS and IBM Modeler software were utilized for data analysis. Results: In 2017, approximately 150 981 752 drug items were issued to outpatients referred to SSO healthcare facilities in Iran. The average number of drug items per outpatient prescription was estimated at 3.33. The proportion of prescriptions that included an injection was 17.5%, and the rate of prescriptions that included an antibiotic was 37.5%. Factors such as patient sex and age, provider specialty, type of facility, and time of outpatient visit were associated with the risk of inappropriate prescriptions. Conclusions: In this study, all drug prescription criteria exceeded the recommended limits set by the World Health Organization. To improve the current prescription patterns throughout the country, it would be beneficial to provide providers with monthly and annual reports and to consider implementing some prescription policies for physicians.

• Journal of Information Processing Systems
• /
• v.20 no.3
• /
• pp.375-390
• /
• 2024

Edge computing architecture has effectively alleviated the computing pressure on cloud platforms, reduced network bandwidth consumption, and improved the quality of service for user experience; however, it has also introduced new security issues. Existing anomaly detection methods in big data scenarios with cloud-edge computing collaboration face several challenges, such as sample imbalance, difficulty in dealing with complex network traffic attacks, and difficulty in effectively training large-scale data or overly complex deep-learning network models. A lightweight deep-learning model was proposed to address these challenges. First, normalization on the user side was used to preprocess the traffic data. On the edge side, a trained Wasserstein generative adversarial network (WGAN) was used to supplement the data samples, which effectively alleviates the imbalance issue of a few types of samples while occupying a small amount of edge-computing resources. Finally, a trained lightweight deep learning network model is deployed on the edge side, and the preprocessed and expanded local data are used to fine-tune the trained model. This ensures that the data of each edge node are more consistent with the local characteristics, effectively improving the system's detection ability. In the designed lightweight deep learning network model, two sets of convolutional pooling layers of convolutional neural networks (CNN) were used to extract spatial features. The bidirectional long short-term memory network (BiLSTM) was used to collect time sequence features, and the weight of traffic features was adjusted through the attention mechanism, improving the model's ability to identify abnormal traffic features. The proposed model was experimentally demonstrated using the NSL-KDD, UNSW-NB15, and CIC-ISD2018 datasets. The accuracies of the proposed model on the three datasets were as high as 0.974, 0.925, and 0.953, respectively, showing superior accuracy to other comparative models. The proposed lightweight deep learning network model has good application prospects for anomaly traffic detection in cloud-edge collaborative computing architectures.

• Tuberculosis and Respiratory Diseases
• /
• v.87 no.4
• /
• pp.473-482
• /
• 2024

Background: Chronic airway diseases, such as asthma and chronic obstructive pulmonary disease (COPD), are increasingly being treated with inhaled corticosteroid (ICS). However, ICSs carry potential infection risks, particularly nontuberculous mycobacteria (NTM). This study investigated the association between ICS use and NTM infection risk using national insurance data, particularly for individuals with chronic airway diseases. Methods: We conducted a nationwide population-based study using data from the National Health Insurance Service-National Sample Cohort in South Korea from 2002 to 2019. The cohort included 57,553 patients diagnosed with COPD or asthma. To assess the risk of NTM infection, we used Cox proportional hazards models and propensity score-based inverse probability of treatment weighting (IPTW) to ensure a balanced analysis of covariates. Results: Of the 57,553 patients (mean age 56.0 years, 43.2% male), 16.5% used ICS and 83.5% did not. We identified 63 NTM infection cases, including nine among ICS users and 54 among non-users. Before and after IPTW, ICS use was associated with a higher risk of NTM infection (adjusted hazard ratio [HR], 4.01; 95% confidence interval [CI], 1.48 to 15.58). Higher risks were significant for patients ≥65 years (adjusted HR, 6.40; 95% CI, 1.28 to 31.94), females (adjusted HR, 10.91; 95% CI, 2.24 to 53.20), never-smokers (adjusted HR, 6.31; 95% CI, 1.49 to 26.64), systemic steroid users (adjusted HR, 50.19; 95% CI, 8.07 to 312.19), and those with higher comorbidity scores (adjusted HR, 6.64; 95% CI, 1.19 to 37.03). Conclusion: ICS use in patients with chronic airway diseases might increase the risk of NTM infection, particularly in older females, never-smokers, and systemic steroid users.

• The Journal of Korea Institute of Information, Electronics, and Communication Technology
• /
• v.17 no.4
• /
• pp.221-228
• /
• 2024

Predicting insurance claims is a key task for insurance companies to manage risks and maintain financial stability. Accurate insurance claim predictions enable insurers to set appropriate premiums, reduce unexpected losses, and improve the quality of customer service. This study aims to enhance the performance of insurance claim prediction models by applying ensemble learning techniques. The predictive performance of models such as Random Forest, Gradient Boosting Machine (GBM), XGBoost, Stacking, and the proposed Dynamic Weighted Ensemble (DWE) model were compared and analyzed. Model performance was evaluated using Mean Absolute Error (MAE), Mean Squared Error (MSE), and the Coefficient of Determination (R²). Experimental results showed that the DWE model outperformed others in terms of evaluation metrics, achieving optimal predictive performance by combining the prediction results of Random Forest, XGBoost, LR, and LightGBM. This study demonstrates that ensemble learning techniques are effective in improving the accuracy of insurance claim predictions and suggests the potential utilization of AI-based predictive models in the insurance industry.

• The Journal of the Convergence on Culture Technology
• /
• v.10 no.5
• /
• pp.673-679
• /
• 2024

Recently, various contents set in the Josun Dynasty have emerged. However, there is a growing number of hanboks that do not preserve the basic original form of hanbok, such as hanboks that deviate significantly from the form of hanboks. Therefore, in this study, a system was created and implemented in the game to express various colors and patterns of hanbok using the DALL-E API in this game by examining the basic hanbok form of hanbok through literature review. However, despite the fact that the results are inconsistent depending on the quality of the Generative AI service and the limitations of not passing through the historical evidence of traditional hanbok experts, it is meaningful to present a system that uses the creativity of users to create a design that can enhance the aesthetics of traditional hanbok. In future research, we want to investigate users' preferences and develop a system that can create an image that fits them when clicking the interface.

• The Journal of Korean Medicine
• /
• v.45 no.3
• /
• pp.131-142
• /
• 2024

Background: The Republic of Korea has maintained a stable number of convalescent hospitals; however, the presence of Korean Medicine(KM) doctors in such establishments has increased. Despite this trend, it is hard to find the recent researches on KM within convalescent hospitals. Objective: This study aimed to investigate the current status of convalescent hospitals, particularly focused on patient characteristics and the provision of KM. Method: Data were obtained from the Health Insurance Review&Assessment Service, relying on the Korea Institute Healthcare Accreditation, spanning from January to March 2024. Analysis was conducted through the categorization of in-patients and the provision of KM, utilizing R software. Results: Among the 143 convalescent hospitals, the majority(83.9%) provided KM, with 43.4% of them located near Seoul. Convalescent hospitals offering KM were characterized by a higher number of doctors(p=0.003) and a greater total bed number(p<0.001). The hospitals with KM specializing in dementia care located near Seoul and exhibited a higher proportion of doctors, total beds(p=0.010), uninsured beds, grade of certification evaluation(Gr) and supported activities of daily living(ADL)(p<0.001, respectively). However, the hospitals with KM specializing in cancer care had higher rate of doctors(p=0.036), total beds, uninsured beds, Gr, and average daily out-of-pocket per person and lower levels of self ADL(p<0.001, respectively). In addition, the hospitals with specialists in KM belonging to dementia care only located near Seoul(p=0.042) and exhibited a higher rate of total beds(p=0.007). Conclusion: These findings indicate significant differences among convalescent hospitals based on patient characteristics and the provision of KM. Consequently, such distinctions merit consideration in future studies.