• 한국정보전자통신기술학회논문지
• /
• 제16권6호
• /
• pp.526-532
• /
• 2023

낙상사고는 세계적으로 매년 42만 건 이상 발생하는 치명적인 사고이다. 따라서, 낙상 환자를 연구하고자 낙상환자의 손상외인코드와 주진단 S코드의 연관성을 찾고, 낙상 환자의 주진단 S코드 데이터를 가지고 손상외인코드를 예측할 수 있는 예측모델을 개발하였다. 본 연구에서는 강원특별자치도 강릉시에 있는 A 기관의 2020~2021년 2년간의 데이터를 받아 낙상에 관련된 손상외인코드 W00~W19까지 데이터만 추출하고, 낙상 손상외인코드 중 예측모형을 개발할 정도의 주진단 S코드를 가지고 있는 W01, W10, W13, W18 데이터를 가지고 예측모형 개발하였다. 데이터 중 80%는 훈련용 데이터, 20%는 테스트용 데이터로 분류하였다. 모형 개발은 MLP(Multi-Layer Perceptron)을 이용하여 6개의 변수(성별, 나이, 주진단S코드, 수술유무, 입원유무, 음주유무)를 입력층에 64개의 노드를 가진 2개의 은닉층, 출력층은 softmax 활성화 함수를 이용하여 손상외인코드 W01, W10, W13, W18 총 4개의 노드를 가진 출력층으로 구성하여 개발하였다. 학습결과 첫 번째 학습했을 때 31.2%의 정확도를 가졌지만, 30번째는 87.5%의 정확도를 나타냈고 이를 통해 낙상환자의 낙상외인코드와 주진단 S코드의 연관성을 확인할 수 있었다.

• Journal of Periodontal and Implant Science
• /
• 제48권2호
• /
• pp.114-123
• /
• 2018

Purpose: The aim of the current study was to develop a computer-assisted detection system based on a deep convolutional neural network (CNN) algorithm and to evaluate the potential usefulness and accuracy of this system for the diagnosis and prediction of periodontally compromised teeth (PCT). Methods: Combining pretrained deep CNN architecture and a self-trained network, periapical radiographic images were used to determine the optimal CNN algorithm and weights. The diagnostic and predictive accuracy, sensitivity, specificity, positive predictive value, negative predictive value, receiver operating characteristic (ROC) curve, area under the ROC curve, confusion matrix, and 95% confidence intervals (CIs) were calculated using our deep CNN algorithm, based on a Keras framework in Python. Results: The periapical radiographic dataset was split into training (n=1,044), validation (n=348), and test (n=348) datasets. With the deep learning algorithm, the diagnostic accuracy for PCT was 81.0% for premolars and 76.7% for molars. Using 64 premolars and 64 molars that were clinically diagnosed as severe PCT, the accuracy of predicting extraction was 82.8% (95% CI, 70.1%-91.2%) for premolars and 73.4% (95% CI, 59.9%-84.0%) for molars. Conclusions: We demonstrated that the deep CNN algorithm was useful for assessing the diagnosis and predictability of PCT. Therefore, with further optimization of the PCT dataset and improvements in the algorithm, a computer-aided detection system can be expected to become an effective and efficient method of diagnosing and predicting PCT.

• 대한수의학회지
• /
• 제55권2호
• /
• pp.89-95
• /
• 2015

Johne's disease, caused by Mycobacterium avium subsp. paratuberculosis (MAP), is one of the most widespread and economically important diseases in cattle. Current diagnostic methods are based on the detection of anti-MAP antibodies in serum or isolation of the causative agent. However, these techniques are often not applicable for cases of subclinical infection due to relatively low sensitivity. Therefore, finding new antigen candidates that strongly react with the host immune system had been attempted. To effectively detect infection during the subclinical stage, several antigen candidates were selected based on previous researches. Characteristics of the selected antigen candidates were analyzed using bioinformatics-based prediction tools. A total of nine antigens were selected (MAP0862, MAP3817c, MAP2077c, MAP0860c, MAP3954, MAP3155c, MAP1204, MAP1087, and MAP2963c) that have MAP-specific and/or high immune responses to infected animals. Using a transmembrane prediction tool, five of the nine antigen candidates were predicted to be membrane protein (MAP3817c, MAP3954, MAP3155c, MAP1087, and MAP1204). Some of the predicted protein structures identified using the I-TASSER server shared similarities with known proteins found in the Protein Data Bank database (MAP0862, MAP1204, and MAP2077c). In future studies, the characteristics and diagnostic efficiency of the selected antigen candidates will be evaluated.

• International journal of advanced smart convergence
• /
• 제10권2호
• /
• pp.145-150
• /
• 2021

Patients with Pressure injuries (PIs) may have pain and discomfort, which results in poorer patient outcomes and additional cost for treatment. This study was a part of larger research project that aimed at prediction modeling using a big data. The purpose of this study were to describe the characteristics of patients with PI in critical care; and to explore comorbidity and diagnostic and interventive procedures that have been done for patients in critical care. This is a secondary data analysis. Data were retrieved from a large clinical database, MIMIC-III Clinical database. The number of unique patients with PI was 2,286 in total. Approximately 60% were male and 68.4% were White. Among the patients, 9.9% were dead. In term of discharge disposition, 56.2% (33.9% Home, 22.3% Home Health Care) where as 32.3% were transferred to another institutions. The rest of them were hospice (0.8%), left against medical advice (0.7%), and others (0.2%). The top three most frequently co-existing kinds of diseases were Hypertension, not otherwise specified (NOS), congestive heart failure NOS, and Acute kidney failure NOS. The number of patients with PI who have one or more procedures was 2,169 (94.9%). The number of unique procedures was 981. The top three most frequent procedures were 'Venous catheterization, not elsewhere classified,' and 'Enteral infusion of concentrated nutritional substances.' Patient with a greater number of comorbid conditions were likely to have longer length of ICU stay (r=.452, p<.001). In addition, patient with a greater number of procedures that were performed during the admission were strongly tend to stay longer in hospital (r=.729, p<.001). Therefore, prospective studies focusing on comorbidity; and diagnostic and preventive procedures are needed in the prediction modeling of pressure injury development in ICU patients.

• 한국구조물진단유지관리공학회 논문집
• /
• 제16권5호
• /
• pp.148-160
• /
• 2012

본 연구에서는 콘크리트 구조물의 합리적인 압축강도 및 탄산화 예측식을 제안하고, 이 제안식의 현장적용을 통한 보다 효율적인 터널 진단 및 유지관리 기법을 개발하였다. 이를 위하여 공용연수가 약 30년 이상 경과하였으며, 약 15년 동안 수회에 걸친 진단 및 점검으로 무수히 많은 현장 내구성 측정 데이터가 축적된 서울메트로를 대상 시설물로 선정하였다. 압축강도 및 탄산화 분석결과 80% 이상의 정확도를 확보하는 각각의 예측식을 도출하였으며, 기존 제안식과의 비교분석을 통하여 본 연구 제안식의 신뢰도를 확인하였다. 또한 제안식의 현장적용 결과 압축강도 및 탄산화 깊이에 대한 예측치의 평균오차율이 약 20%내외로서 80% 이상의 높은 정확도를 확보하는 것으로 분석되어 현장적용의 적합성을 확인하였다. 현장조사 전 내구성 예측 맵(Map)을 활용한 효율적인 유지관리 기법을 개발하였다. 예측 맵(Map) 활용 시 진단기술자 및 시설물 담당자는 설계기준강도에 미달되거나 탄산화로 철근부식 가능성이 높은 취약부위를 한 눈에 파악할 수 있으므로 일일이 조사를 수행하는 과정에서 취약부위를 도출해야 하는 현 조사기법 보다 효과적으로 터널 조사 및 유지관리를 수행할 수 있을 것으로 기대된다.

• 한국대기환경학회지
• /
• 제17권1호
• /
• pp.1-17
• /
• 2001

The high-ozone episode in the Greater Seoul Area for the period of July 27 to August 1 1997 was modeled by the CIT(California Institute of Technology) three-dimensional photochemical model. Emission data were prepared by scaling the NIER(1994) data through and optimization method using VOC measurements in August 1997 and EKMA(Empirical Kinetic Modeling Approach). Two sets of meteorological data were prepared by the diagnostic routine. a part of the CIT model : one only utilized observations from the surface weather stations and the other also utilized observations from the automatic weather stations that were more densely distributed than those from the surface weather stations. The results showed that utilizing observations from the automatic weather stations could represent fine variations in the sind field such as those caused by topography. A better wind field gave better peak ozones and a more reasonable spatial distribution of ozone concentrations. Nevertheless, there were still many differences between predictions and observations particularly for primary pollutant such as NOx and CO. This was probably due to the inaccuracy of emission data that could not resolve both temporal and spatial variations.

• 응용통계연구
• /
• 제21권1호
• /
• pp.109-123
• /
• 2008

많은 소지역 추정량이 제안되었으며, 국내외에서 소지역 추정에 관한 많은 연구가 진행되고 있다. 또한 소지역 추정량의 특성과 우수성을 비교하기위한 비교통계량도 연구되고 있다. 기존의 소지역 추정량은 MSE(Mean square error)를 최소화하여 얻어지며, 이에 따라 추정량의 우수성도 MSE를 기준으로 판단하고 있다. 본 논문에서는 최근 새롭게 재조명 되고 있는 MSPE(Mean square percentage error)를 최소화하는 추정량을 제안하였다. 신기일 등 (2007)에서 사용된 비교통계량과 MSE 그리고 MSPB를 이용하여 제안된 추정량과 기존의 소지역 추정량을 비교하였다.

• 한국표면공학회지
• /
• 제43권3호
• /
• pp.159-164
• /
• 2010

Surface roughness of deposited or etched film strongly depends on ion bombardment. Relationships between ion bombardment variables and surface roughness are too complicated to model analytically. To overcome this, an empirical neural network model was constructed and applied to a deposition process of silicon nitride (SiN) films. The films were deposited by using a pulsed plasma enhanced chemical vapor deposition system in $SiH_4$-$NH_4$ plasma. Radio frequency source power and duty ratio were varied in the range of 200-800 W and 40-100%. A total of 20 experiments were conducted. A non-invasive ion energy analyzer was used to collect ion energy distribution. The diagnostic variables examined include high (or) low ion energy and high (or low) ion energy flux. Mean surface roughness was measured by using atomic force microscopy. A neural network model relating the diagnostic variables to the surface roughness was constructed and its prediction performance was optimized by using a genetic algorithm. The optimized model yielded an improved performance of about 58% over statistical regression model. The model revealed very interesting features useful for optimization of surface roughness. This includes a reduction in surface roughness either by an increase in ion energy flux at lower ion energy or by an increase in higher ion energy at lower ion energy flux.

• 한국전자통신학회논문지
• /
• 제17권5호
• /
• pp.825-832
• /
• 2022

본 논문은 영농형 태양광 발전 시스템의 전력 생산량을 수집·저장하여 지능적인 예측 모델을 구현하기 위한 예측 및 진단 모델의 설계와 구현에 대해 논한다. 제안된 모델은 시계열 데이터에 특화된 순환신경망 기법인 RNN, LSTM, GRU 모델을 이용하여 태양광 발전량을 예측하고 각 모델의 하이퍼 파라미터를 다르게 주어 비교 분석하고, 성능을 평가했다. 그 결과 세 모델 모두 MSE, RMSE 지표는 0에 매우 가까우며, R2 지표는 1에 가까운 성능을 보였다. 이를 통해 제안하는 예측 모델은 태양광 발전량을 예측하기에 적합한 모델임을 알 수 있고, 이러한 예측을 이용하여 영농형 태양광 시스템에서 지능적인 운영관리 기능에 적용될 수 있음을 보였다.

• Asian Pacific Journal of Cancer Prevention
• /
• 제17권9호
• /
• pp.4377-4380
• /
• 2016

Background: To compare diagnostic performance in differentiating benign from malignant ovarian masses between IOTA (the International Ovarian Tumor Analysis) simple rules and subjective sonographic assessment. Materials and Methods: Women scheduled for elective surgery because of ovarian masses were recruited into the study and underwent ultrasound examination within 24 hours of surgery to apply the IOTA simple rules by general gynecologists and to record video clips for subjective assessment by an experienced sonographer. The diagnostic performance of the IOTA rules and subjective assessment for differentiation between benign and malignant masses was compared. The gold standard diagnosis was pathological or operative findings. Results: A total of 150 ovarian masses were covered, comprising 105 (70%) benign and 45 (30%) malignant. Of them, the IOTA simple rules could be applied in 119 (79.3%) and were inconclusive in 31 (20.7%) whereas subjective assessment could be applied in all cases (100%). The sensitivity and the specificity of the IOTA simple rules and subjective assessment were not significantly different, 82.9% vs 86.7% and 94.0% vs 94.3% respectively. The agreement of the two methods in prediction was high with a Kappa index of 0.835. Conclusions: Both techniques had a high diagnostic performance in differentiation between benign and malignant ovarian masses but the IOTA rules had a relatively high rate of inconclusive results. The IOTA rules can be used as an effective screening technique by general gynecologists but when the results are inconclusive they should consult experienced sonographers.