http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
이재준 ( Jaejun Lee ),김태윤 ( Taeyun Kim ),이승우 ( Seong-woo Lee ),김명중 ( Myoung Jung Kim ),한창인 ( Chang In Han ),신유나 ( Yu Na Shin ),김미영 ( Mi Young Kim ),정숙영 ( Suk Young Jung ) 국군의무사령부 2023 대한군진의학학술지 Vol.54 No.1
Objective Prevalence of NAFLD is increasing globally. It is anticipated to be around 16% among young men in Korea, and those with NAFLD might be considered for active surveillance. Moreover, NAFLD with fibrotic burden has a higher risk of liver related events. The purpose of this study was to identify factors associated with liver fibrosis in young NAFLD soldiers and establish a scoring system using these factors. Method We conducted a cross-sectional study between July 2022 and January 2023, enrolling patients with NAFLD who presented with elevated levels of alanine transaminase. The FAST score was adopted to determine the presence of significant fibrosis. FAST scores greater than 0.67 were considered to indicate significant fibrosis. To develop a novel scoring system for predicting significant fibrosis, participants were randomly assigned to the derivation and the validation cohort. Results A total of 436 patients were enrolled and 67 patients were assessed to have significant fibrosis by the FAST score, indicating a 15.4% prevalence of significant fibrosis. Variables such as ALT levels, BMI, and hypertriglyceridemia were found to be associated with significant fibrosis. Two models were derived from the derivation cohort to predict significant fibrosis. The models were validated using the validation cohort and showed excellent performance (Model 1: AUROC=0.924, Model 2: AUROC=0.894). Conclusion This study identified ALT levels, BMI, and hypertriglyceridemia as factors associated with significant fibrosis in young soldiers with NAFLD. Two models with excellent performance were developed using these variables to predict significant fibrosis. However, biopsy-proven data are required to validate the performance of the proposed scoring system.
Lee Jaejun,Jung Jae Hyeop,Choi Sung Jun,Ha Beomman,Yang Hyun,Sung Pil Soo,Bae Si Hyun,Yu Jeong-A 거트앤리버 소화기연관학회협의회 2024 Gut and Liver Vol.18 No.5
Background/Aims: Young Korean men are obligated to serve in the military for 18 to 21 months. We investigated the effects of military service on steatotic liver disease (SLD) and other metabolic parameters. Methods: Pre-enlistment health check-up performed from 2019 to 2022 and in-service health check-up performed from 2020 to 2022 were merged as paired data. SLD was defined as a hepatic steatosis index of 36 or higher. Hypertension (HTN) and hypertriglyceridemia were also included in the analysis. Results: A total of 503,136 paired cases were included in the analysis. Comparing pre-enlistment and in-service health check-ups, the prevalence of SLD (22.2% vs 17.6%, p<0.001), HTN (7.6% vs 4.3%, p<0.001), and hypertriglyceridemia (8.1% vs 2.9%, p<0.001) decreased during military service. In terms of body mass index, the proportion of underweight (8.2% vs 1.4%, p<0.001) and severely obese (6.1% vs 4.9%, p<0.001) individuals decreased over time. Regarding factors associated with SLD development and resolution, age was positively associated with SLD development (odds ratio, 1.146; p<0.001) and a health check-up interval of <450 days was a protective factor for SLD development (odds ratio, 0.746; p<0.001). Those serving in the marines were less likely to develop SLD, whereas those serving in the navy were more likely to develop SLD. Serving in the army or the navy was negatively associated with SLD resolution, whereas serving in the air force was positively associated with SLD resolution. Conclusions: The prevalence of SLD, HTN, and hypertriglyceridemia decreased substantially during Korean military service.
이재준(Jaejun Lee),서호건(Hogeon Seo),유용균(Yonggyun Yu) 한국비파괴검사학회 2024 한국비파괴검사학회지 Vol.44 No.3
본 연구는 다채널 센서 신호에서 특정 채널이나 값이 소실되었을 때, 오토인코더 모델을 활용하여 소실된 값을 복원하는 방법을 제안한다. 이를 위해 정상데이터를 사용해 오토인코더 모델을 학습시킨 후 소실된 채널의 값을 예측하였다. 평가 결과, 소실되지 않은 채널의 정보를 활용하여 원래의 값과 패턴에 근사하게 복원이 이루어졌으며, 채널 간 상관관계에 따라 복원 성능이 달라질 수 있음을 확인하였다. 제안한 방법을 통해 전체 데이터의 유효성을 향상시킬 수 있으며, 센서 고장이나 데이터 소실 상황에서 데이터 복원력을 높이는 데 기여할 수 있다. We propose a method for restoring lost values in multi-channel sensor signals when specific channels or values are missing by using an autoencoder model. For this purpose, an autoencoder model was trained using normal data and then used to predict the values of the missing channels. Evaluation results showed that the restoration approximated the original values and patterns by utilizing information from the non-missing channels. Additionally, the restoration performance varied, depending on the correlations among different channels. The proposed method can enhance the overall validity of a dataset and contribute to the improvement of the data restoration capability in situations of sensor failures or data loss.