http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
유승목,Pai H. Chou 한국전자통신연구원 2012 ETRI Journal Vol.34 No.4
Serial peripheral interface (SPI) has been identified as a bottleneck in many wireless sensing systems today. SPI is used almost universally as the physical connection between the microcontroller unit (MCU) and radios, storage devices, and many types of sensors. Virtually all wireless sensor nodes today perform up to twice as many bus transactions as necessary to transfer a given piece of data, as an MCU must serve as the bus master in all transactions. To eliminate this bottleneck, we propose the master-handoff protocol. After the MCU initiates reading from the source slave device and writing to the sink slave device, the MCU as a master becomes a slave, and either the source or the sink slave becomes the temporary master. Experiment results show that this master-handoff technique not only cuts the data transfer time in half, but, more importantly, also enables a superlinear energy reduction.
유승목,이두호,김종기,소찬영,홍준표 ( Seung Mog Yoo,Doo Ho Lee,Jong Ki Kim,Chan Young So,Chun Pyo Hong ) 한국주조공학회 1991 한국주조공학회지 Vol.11 No.1
N/A A computer program which consists of pre-processor, main solidification simulator and post-processor has been developed for three dimensional solidification analysis of steel castings. The pre-processor is used for mesh generation in a small personal computing system. The modified finite difference method is adopted for the main solidification simulation algorithm. The post -processor graphically presents the simulation results and shows the formation of shrinkage defects. Several experiments on large steel castings in sand mold were carried out. The temperature variations in casting and mold with time are measured experimentally, and the results are compared with calculation results. Several numerical examples for the prediction of shrinkage cavity in large steel casting of SC42 and SCNCrM2 alloys are compared with experimental results. The effect of sleeve and chills on solidification patterns are also studied. Formation of shrinkage defects for the three cases of experimental castings are relatively well predicted by present model.
패혈증성 쇼크환자에서 지속적 신대체요법 시작시기와 예후와의 관계
유승목,김원영,최상식,허진원,홍상범,임채만,고윤석 대한중환자의학회 2012 Acute and Critical Care Vol.27 No.1
Background: Initiation of renal replacement therapy (RRT) in critically ill septic shock patients with acute kidney injury is highly subjective and may influence outcome. The aim of this study is to evaluate the relationship between initiation of RRT and 28 day mortality in patients with severe sepsis and septic shock (SSSS). Methods: All patients diagnosed with SSSS and treated at the medical intensive care unit (ICU) in university-affiliated hospital from January 2005 to December 2006 were reviewed. Initiation of RRT was stratified into “early” and “late” by RIFLE (Risk, Injury, Failure, Loss, and End-stage) criteria and blood urea nitrogen (BUN) at the time RRT began. The primary outcome was death after 28 days from any cause. Results: Of the 326 patients diagnosed with SSSS and admitted into the medical ICU during the study period, 78 patients received RRT. Mean age was 61.5 ± 14.7 years old and 54 patients were male (69.2%). The initiation of RRT was categorized into early (Risk, and Injury) and late (Failure) by RIFLE criteria and also categorized into early (BUN < 75 mg/dl) and late (BUN ≥ 75 mg/dl). When the relationship between RIFLE criteria and 28 day mortality was compared, no significant difference was shown (70.8% vs. 73.3%, p = 0.81). The initiation of RRT by BUN also showed no significant difference in 28 day mortality (77.3% vs. 69.6%, p = 0.50). Conclusions: Initiation of RRT, stratified into “early” and “late” by RIFLE and BUN, showed no significant difference in 28 day mortality regarding patient with SSSS.
유승목,Gina Yu,Seung Joon Yoo,Sang-Hun Lee,June Sung Kim,Sungmin Jung,Youn-Jung Kim,김원영 대한중환자의학회 2019 Acute and Critical Care Vol.34 No.2
Background: The current Surviving Sepsis Campaign guidelines recommend the remeasure ment of lactate levels if the initial lactate level is elevated; however, the prognostic value of lactate kinetics is limited and inconsistent. We attempted to determine the efficacy of the lactate area score (calculated from repeated lactate measurements during initial resuscitation) as a prognostic marker of septic shock in the emergency department (ED). Methods: We performed a retrospective study of adult patients with septic shock in the ED of a single tertiary medical center. Serial lactate levels were measured five times within 12 hours. We also compared the initial lactate level, maximum lactate level, and lactate area score. The lactate area score was defined as the sum of the area under the curve measured at 2, 4, 6, and 12 hours following the initial measurement. Results: A total of 362 patients were enrolled in this study, and the overall 28-day mortality was 31.8%. The lactate area score of serial lactate levels as well as the initial (median [inter quartile range], 4.9 [3.4 to 10.5]; P =0.003) and maximum (7.3 [4.2 to 13.2]; P<0.001) lactate levels were significantly higher in the non-survivor group. However, in multivariate analysis, only the lactate area score (odds ratio, 1.013; 95% confidence interval, 1.007 to 1.019) was significantly associated with 28-day mortality. Conclusions: The early lactate area score may be a possible prognostic marker for predicting the 28-day mortality of adult septic shock patients. Further prospective interventional studies should be conducted to validate our results.
유승목,이경희,박재복,윤석진,조창식,정영준,조일연,Yoo, Seung-mok,Lee, Kyung Hee,Park, Jaebok,Yoon, Seok Jin,Cho, Changsik,Jung, Yung Joon,Cho, Il Yeon 한국전자통신연구원 2019 전자통신동향분석 Vol.34 No.4
Deep learning is a hot topic in both academic and industrial fields. Deep learning applications can be categorized into two areas. The first category involves applications such as Google Alpha Go using interfaces with human operators to run complicated inference engines in high-performance servers. The second category includes embedded applications for mobile Internet-of-Things devices, automotive vehicles, etc. Owing to the characteristics of the deployment environment, applications in the second category should be bounded by certain H/W and S/W restrictions depending on their running environment. For example, image recognition in an autonomous vehicle requires low latency, while that on a mobile device requires low power consumption. In this paper, we describe issues faced by embedded applications and review popular inference engines. We also introduce a project that is being development to satisfy the H/W and S/W requirements.