New motor development requires high-speed load testing using dynamo equipment to calculate the efficiency of the motor. Abnormal noise and vibration may occur in the test equipment rotating at high speed due to misalignment of the connecting shaft or looseness of the fixation, which may lead to safety accidents. In this study, three single-axis vibration sensors for X, Y, and Z axes were attached on the surface of the test motor to measure the vibration value of vibration. Analog data collected from these sensors was used in classification models for anomaly detection. Since the classification accuracy was around only 93%, commonly used hyperparameter optimization techniques such as Grid search, Random search, and Bayesian Optimization were applied to increase accuracy. In addition, Response Surface Method based on Design of Experiment was also used for hyperparameter optimization. However, it was found that there were limits to improving accuracy with these methods. The reason is that the sampling data from an analog signal does not reflect the patterns hidden in the signal. Therefore, in order to find pattern information of the sampling data, we obtained descriptive statistics such as mean, variance, skewness, kurtosis, and percentiles of the analog data, and applied them to the classification models. Classification models using descriptive statistics showed excellent performance improvement. The developed model can be used as a monitoring system that detects abnormal conditions of the motor test.

Anomaly detection of Machine Learning such as PCA anomaly detection and CNN image classification has been focused on cross-sectional data. In this paper, two approaches has been suggested to apply ML techniques for identifying the failure time of big time series data. PCA anomaly detection to identify time rows as normal or abnormal was suggested by converting subjects identification problem to time domain. CNN image classification was suggested to identify the failure time by re-structuring of time series data, which computed the correlation matrix of one minute data and converted to tiff image format. Also, LASSO, one of feature selection methods, was applied to select the most affecting variables which could identify the failure status. For the empirical study, time series data was collected in seconds from a power generator of 214 components for 25 minutes including 20 minutes before the failure time. The failure time was predicted and detected 9 minutes 17 seconds before the failure time by PCA anomaly detection, but was not detected by the combination of LASSO and PCA because the target variable was binary variable which was assigned on the base of the failure time. CNN image classification with the train data of 10 normal status image and 5 failure status images detected just one minute before.

목 적 : 복부 dynamic검사 시 조영제 주입 후 환자의 이상 반응 발생 빈도 및 조영제의 신호 대 잡음비 분석을 통해 적정한 조영제 주입 속도를 알아보고자 하였다. 대상 및 방법 : 2013년 4월부터 2014년 9월까지 3.0T MRI system(Magnetom Tim Trio, Siemens, Germany)을 이용하여 Gadoxetic acid(Primovist, Bayer Schering Pharma)를 사용한 309명(남자 223명, 여자 86명)을 대상으로 Liver dynamic검사를 하였고, Gadodiamide(Ominisacn, GE healthcare)을 사용한 247명(남자 148명, 여자 99명)을 대상으로 Cholangio, Pancreas MRI 상복부 검사를 시행하였다. Saline은 주입 속도를 조영제와 동일하게 주입했으며 그 양은 30ml로 고정하고 gadoxetic acid의 양은 0.1ml/kg, 조영제의 주입속도는 1.0cc/sec, 1.5cc/sec, 2.0cc/sec로 60sec, 100sec을 측정했다. Gadodiamide의 양은 0.2ml/kg, 조영제 주입 속도는 1.5cc/sec, 2.0cc/sec, 2.5cc/sec로 60sec, 120sec을 측정했으며 Volumetric Interpolated Breath-hold Examination(VIBE) sequence를 적용하여 영상을 획득하였다. 상복부 MRI검사 시 조영제의 주입속도에 따른 이상 반응 유무 및 발생 빈도를 분석하고 획득한 영상은 PACS viewer에서 각 phase 별로 ROI를 5~15mm로 main portal vein, liver parenchyma, aorta를 세 곳 설정하여 각 부위의 SNR을 60초, 100초, 120초 검사 중심으로 정량 분석하였다. 결 과 : Liver MRI의 arterial phase 시간대 이상 반응(일시적 호흡곤란, 기침, 구토, 가슴 답답함 등) 발생 빈도는 2.0cc/sec는 98명 중 23건(23.5%), 1.5cc/sec는 116명 중 44건(37.9%), 1.0cc/sec에서 95명 중 22건(23.2%)으로 나타났다. SNR은 조영제주입 속도를 2.0cc/sec로 했을 때 60초 검사에서 portal vein 439.13±38.26, liver parenchyma 250.54±28.54, aorta 311.74±33.24였고 100초에서는 312.01±20.29, 283.88±28.60, 250.38±22.68이었다. 1.5cc/sec의 경우 60초 검사에서 portal vein 415.81±42.43, parenchyma 247.14±28.24, aorta 305.48±30.16, 100초엔 각각 339.20±32.12, 273.65±29.28, 264.66±29.84였다. 1.0cc/sec의 경우 60초 검사는 portal vein 400.44±42.43, parenchyma 237.95±28.30, aorta 307.24±33.69, 100초는 359.17±31.67, 259.13±27.90, 280.49±32.74였다. Cholangio MRI는 2.5cc/sec는 90명 중 6명(6.7%), 2.0cc/sec는 87명 중 3명(3.4%), 1.5cc/sec에서 70명 중 4명(5.7%)이 arterial phase에서 이상 반응이 발생됐다. 신호대 잡음비는 2.5cc/sec로 조영제 주입 시 60초 검사에서 portal vein 659.09±59.10, liver parenchyma 367.40±39.09, aorta 537.40±41.82였으며 120초에서 각각 546.79±40.35, 376.15±28.79, 470.59±36.99였다. 2.0cc/sec로 주입했을 땐 60초에 portal vein 648.18±61.19, parenchyma 360.49±41.35, aorta 540.07±51.77이었고 120초는 각각 559.26±52.35, 343.70±38.27, 474.57±42.02였으며 1.5cc/sec로 주입 시엔 60초에 portal vein 637.62±54.24, parenchyma 358.85±42.87, aorta 544.25±47.16였고 120초는 562.35±39.57, 303.91±40.20, 481.22±40.14이었다. 결 론 : 조영제의 주입 속도를 2.0cc/sec로 했을 경우 Liver MRI는 저속 주입 시 보다 60초와 100초 검사에서 이상 반응의 증가없이 SNR이 더 좋은 영상을 획득했다. Cholangio MRI 또한 60초와 120초 검사에서 2.5cc/sec로 주입했을 때 저속 주입 시 보다 SNR이 더 우수한 영상을 얻을 수 있었다.

Oral lichen planus (OLP) is an atypically keratinized and ulcerative lesion, producing severe pain and discomforts in the involved patients. Nevertheless, the etiological factor or the pathogenetic mechanism has not been clearly elucidated. In the present study, the different gene expressions were screened in 21 cases of OLP by immunohistochemical (IHC) array method using 80 antibodies, and found that the pathway of E-cadherin/β-catenin was abnormally expressed compared to the other essential genetic pathways. Particularly, the expressions of eIF5A, DHS, and DOHH, which are the biomarkers of protein translation, were remarkably reduced, nevertheless the expression of β-catenin was strongly positive in the 7 cases among 21 cases of OLP. The other expressions of p53, BCL-2, MDM-2, PAKT, BAX, BAK, BAD, NFkB, HO-1, etc, were usually weak or sparse, while the expressions of PCNA, CDK4, and HSP-70 were markedly increased. Taken together, it is presumed that the overexpression of β-catenin indicates the derangement of E-cad/β-catenin/NFkB pathway, causing the transcription of cellular proliferating genes in downstream events, i.e., PCNA and CDK4, and that it may be eventually relevant to the malignant potential of OLP epithelial cells. It is also suggested that the activation of β-catenin/TCF/LEF1 pathway be closely relevant to the immunological reaction of OLP with the accumulation of T-cells underneath the mucosal epithelium.