• 한국구조물진단유지관리공학회 논문집
• /
• 제23권3호
• /
• pp.111-118
• /
• 2019

본 연구에서는 비정질강섬유의 혼입이 초고강도콘크리트의 폭렬특성에 미치는 영향이 실험적으로 검토되었다. 콘크리트는 압축강도 100과 150 MPa의 초고강도콘크리트가 사용되었다. 폴리프로필렌섬유는 0.15 vol%, 비정질강섬유는 0.3 및 0.5 vol%가 혼입되었다. 시험체는 콘크리트의 압축강도와 섬유혼입 조건에 따라 6수준이 제작되었고, ISO-834 가열곡선에 의해 가열되었다. 결과로써 폴리프로필렌섬유와 비정질강섬유가 혼입된 초고강도콘크리트의 폭렬제어에 있어서는 용융된 폴리프로필렌섬유가 형성하는 공극네트워크를 통해 수증기가 이동하는 효과가 지배적인 것으로 나타났다. 또한, 비정질강섬유 0.3v ol% 혼입률에서는 폭렬제어에 큰 영향을 미치지 않지만, 0.5 vol%의 비정질강섬유가 혼입될 경우에는 수증기가 이동할 수 있는 균열의 발생이 억제됨으로써 콘크리트 폭렬의 원인으로 지적되고 있는 수분막힘층(moisture clog)가 형성될 가능성이 높은 것을 확인할 수 있었다.

• 정보보호학회논문지
• /
• 제20권1호
• /
• pp.3-16
• /
• 2010

최근 무선 센서 네트워크 보안 분야에서는 키 교환을 위한 부가적인 통신이 필요 없이 통신 엔터티 상호간에 암호화를 수행할 수 있는 페어링 암호가 주목받고 있다. 본 논문에서는 이러한 페어링 암호의 한 종류인 ${\eta}_T$ 페어링에 대한 효율적인 하드웨어 구현을 제시한다. 이를 위해 병렬 처리 및 레지스터/자원의 최적화에 기반한 ${\eta}_T$ 페어링 알고리즘에 대한 효율적인 하드웨어 구조를 제안하며, 제안한 구조를 GF($2^{239}$) 상에서 FPGA로 구현한 결과를 나타낸다. 제안한 구조는 기존의 구현 결과에 비해 Area Time Product에 있어 15% 나은 결과를 가진다.

• 대한방사선기술학회지:방사선기술과학
• /
• 제45권2호
• /
• pp.111-116
• /
• 2022

In this study, the applicability of machine learning for the development of a simulator for general X-ray examination education is evaluated. To this end, k-nearest neighbor(kNN), support vector machine(SVM) and neural network(NN) classification models are analyzed to present the most suitable model by analyzing the results. Image data was obtained by taking 100 photos each corresponding to Posterior anterior(PA), Posterior anterior oblique(Obl), Lateral(Lat), Fan lateral(Fan lat). 70% of the acquired 400 image data were used as training sets for learning machine learning models and 30% were used as test sets for evaluation. and prediction model was constructed for right-handed PA, Obl, Lat, Fan lat image classification. Based on the data set, after constructing the classification model using the kNN, SVM, and NN models, each model was compared through an error matrix. As a result of the evaluation, the accuracy of kNN was 0.967 area under curve(AUC) was 0.993, and the accuracy of SVM was 0.992 AUC was 1.000. The accuracy of NN was 0.992 and AUC was 0.999, which was slightly lower in kNN, but all three models recorded high accuracy and AUC. In this study, right-handed PA, Obl, Lat, Fan lat images were classified and predicted using the machine learning classification models, kNN, SVM, and NN models. The prediction showed that SVM and NN were the same at 0.992, and AUC was similar at 1.000 and 0.999, indicating that both models showed high predictive power and were applicable to educational simulators.

• 대한치과교정학회지
• /
• 제52권4호
• /
• pp.268-277
• /
• 2022

Objective: Poor experience with Invisalign treatment affects patient compliance and, thus, treatment outcome. Knowing the potential discomfort level in advance can help orthodontists better prepare the patient to overcome the difficult stage. This study aimed to construct artificial neural networks (ANNs) to predict patient experience in the early stages of Invisalign treatment. Methods: In total, 196 patients were enrolled. Data collection included questionnaires on pain, anxiety, and quality of life (QoL). A four-layer fully connected multilayer perception with three backpropagations was constructed to predict patient experience of the treatment. The input data comprised 17 clinical features. The partial derivative method was used to calculate the relative contributions of each input in the ANNs. Results: The predictive success rates for pain, anxiety, and QoL were 87.7%, 93.4%, and 92.4%, respectively. ANNs for predicting pain, anxiety, and QoL yielded areas under the curve of 0.963, 0.992, and 0.982, respectively. The number of teeth with lingual attachments was the most important factor affecting the outcome of negative experience, followed by the number of lingual buttons and upper incisors with attachments. Conclusions: The constructed ANNs in this preliminary study show good accuracy in predicting patient experience (i.e., pain, anxiety, and QoL) of Invisalign treatment. Artificial intelligence system developed for predicting patient comfort has potential for clinical application to enhance patient compliance.

• 스마트미디어저널
• /
• 제12권2호
• /
• pp.76-82
• /
• 2023

제4차 산업혁명시대에 P2P 네트워크를 통해 데이터를 분산하여 관리하는 기술인 블록체인은 제조, 문화, 공공 분야 등 다양한 분야에서 탈중앙형의 새로운 네트워킹 패러다임으로써 활용되고 있다. 하지만, 양자 컴퓨터의 등장과 함께 해시 함수, 대칭키 암호, 공개키 암호 등 기존 암호 체계의 문제를 해결할 수 있는 양자 알고리즘이 소개가 되었다. 현재 주요 블록체인 시스템은 대부분 트랜잭션 서명에 타원곡선 암호를 사용하고 있어 양자 공격자로부터 안전하지 않다. 따라서, 블록체인에서 트랜잭션 서명을 위해 격자 기반 암호를 활용하는 양자 내성 블록체인에 대한 연구가 필요하다. 본 논문에서는 양자 내성을 갖는 격자 기반 암호를 활용하여 서명할 내용을 숨겨 서명할 뿐만 아니라, 추후 서명 내용이 검증 가능한 블록체인을 위한 블라인드 서명 기법을 제안한다. 또한, 랜덤 오라클 모델을 이용하여 제안한 기법의 보안성을 검증한다.

• 대한의용생체공학회:의공학회지
• /
• 제44권2호
• /
• pp.139-146
• /
• 2023

Osteoporosis is a disease in which the risk of bone fractures increases due to a decrease in bone density caused by aging. Osteoporosis is diagnosed by measuring bone density in the total hip, femoral neck, and lumbar spine. To accurately measure bone density in the lumbar spine, the vertebral region must be segmented from the lumbar X-ray image. Deep learning-based automatic spinal segmentation methods can provide fast and precise information about the vertebral region. In this study, we used 695 lumbar spine images as training and test datasets for a deep learning segmentation model. We proposed a lumbar automatic segmentation model, CM-Net, which combines the center point of the spine and the modified U-Net network. As a result, the average Dice Similarity Coefficient(DSC) was 0.974, precision was 0.916, recall was 0.906, accuracy was 0.998, and Area under the Precision-Recall Curve (AUPRC) was 0.912. This study demonstrates a high-performance automatic segmentation model for lumbar X-ray images, which overcomes noise such as spinal fractures and implants. Furthermore, we can perform accurate measurement of bone density on lumbar X-ray images using an automatic segmentation methodology for the spine, which can prevent the risk of compression fractures at an early stage and improve the accuracy and efficiency of osteoporosis diagnosis.

• 정보보호학회논문지
• /
• 제33권4호
• /
• pp.687-697
• /
• 2023

공격 양상이 더욱 지능화되고 다양해진 봇넷은 오늘날 가장 심각한 사이버 보안 위협 중 하나로 인식된다. 본 논문은 UGR과 CTU-13 데이터 셋을 대상으로 반지도 학습 딥러닝 모델인 오토엔코더를 활용한 봇넷 검출 실험결과를 재검토한다. 오토엔코더의 입력벡터를 준비하기 위해, 발신지 IP 주소를 기준으로 넷플로우 레코드를 슬라이딩 윈도우 기반으로 그룹화하고 이들을 중첩하여 트래픽 속성을 추출한 데이터 포인트를 생성하였다. 특히, 본 논문에서는 동일한 흐름-차수(flow-degree)를 가진 데이터 포인트 수가 이들 데이터 포인트에 중첩된 넷플로우 레코드 수에 비례하는 멱법칙(power-law) 특징을 발견하고 실제 데이터 셋을 대상으로 97% 이상의 상관계수를 제공하는 것으로 조사되었다. 또한 이러한 멱법칙 성질은 오토엔코더의 학습에 중요한 영향을 미치고 결과적으로 봇넷 검출 성능에 영향을 주게 된다. 한편 수신자조작특성(ROC)의 곡선아래면적(AUC) 값을 사용해 오토엔코더의 성능을 검증하였다.

• 천문학회보
• /
• 제46권1호
• /
• pp.52.1-52.1
• /
• 2021

We present the results of multi-epoch, multi-frequency monitoring of a blazar 4C +29.45, which was regularly monitored as part of the Interferometric Monitoring of GAmma-ray Bright AGNs program - a key science program of the Korean Very long baseline interferometry Network (KVN). Observations were conducted simultaneously at 22, 43, 86 and 129 GHz during the 4 years from December 2012 to December 2016. We also used additional data from the 15 GHz Owens Valley Radio Observatory (OVRO) monitoring program. From the 15 GHz light curve, we estimated the variability time scales of the source during several radio flux enhancements. We found that the source experiencesd 6 radio flux enhancements with variability time scales of 9-187 days during the observing period, yielding corresponding variability Doppler factors of 9-27. From the multi-frequency simultaneous KVN observations, we were able to obtain accurate radio spectra of the source and hence to more precisely measure the turnover frequencies 𝜈_r of synchrotron self-absorbed (SSA) emission with a mean value of ${\bar{\nu}_r}=28.9GHz$. Using jet geometry assumptions, we estimated the size of the emitting region at the turnover frequency. Taking into account these results, we found that the equipartition magnetic field strength is up to two orders of magnitudes higher than the SSA magnetic field strength (0.6-99 mG). This is consistent with the source being particle dominated.

• 천문학회보
• /
• 제46권2호
• /
• pp.49.1-49.1
• /
• 2021

We carried out simultaneous monitoring observations of five maser lines, H₂O (22 GHz), SiO 𝝊 =1, 2, J =1-0 (43.1, 42.8 GHz), and SiO 𝝊 =1, J=2-1, J =3-2 (86.2, 129.3 GHz), toward the Mira variable star WX Serpentis with the 21-m antennas of the Korean VLBI Network (KVN) in 2009-2021 (~12 years). Most spectra of the H₂O maser are well separated into two parts of two blue- and one redshifted features within ± 10 km s-1 of the stellar velocity. All detected SiO masers are generally concentrated within ± 5 km s-1 of the stellar velocity, and sometimes appear split into two components. Overall, the profiles of SiO and H₂O masers detected in WX Serpentis illustrate typical characteristics of the Mira variable. In addition, flux variations of both SiO and H₂O masers are well correlated with the optical light curve of the central star, showing a phase lag of ~ 0.1 for SiO masers and ~ 0.2 for H₂O maser. This phenomenon is considered to be the direct effect of propagating shock waves generated by the stellar pulsation, because SiO and H₂O masers are sequentially distributed at different positions with respect to the central star. In addition, we analyzed long-term trends and characteristics of maser velocities, maser ratio, and the velocity extents (the full width at zero power; FWZP). We also investigated a spectral energy distribution (SED) ranging from 1.2 to 240 ㎛ obtained using several infrared data: 2MASS, WISE, IRAS, ISO, COBE DIBRE, RAFGL, and AKARI (IRC and FIS). From the IRAS LRS and ISO SWS spectra of this star, we identified 9.7 and 12 ㎛ silicate emission features consistent with the SE6 spectrum model, corresponding to the typical AGB phase.

• Tuberculosis and Respiratory Diseases
• /
• 제86권3호
• /
• pp.226-233
• /
• 2023

Background: Inactive or old, healed tuberculosis (TB) on chest radiograph (CR) is often found in high TB incidence countries, and to avoid unnecessary evaluation and medication, differentiation from active TB is important. This study develops a deep learning (DL) model to estimate activity in a single chest radiographic analysis. Methods: A total of 3,824 active TB CRs from 511 individuals and 2,277 inactive TB CRs from 558 individuals were retrospectively collected. A pretrained convolutional neural network was fine-tuned to classify active and inactive TB. The model was pretrained with 8,964 pneumonia and 8,525 normal cases from the National Institute of Health (NIH) dataset. During the pretraining phase, the DL model learns the following tasks: pneumonia vs. normal, pneumonia vs. active TB, and active TB vs. normal. The performance of the DL model was validated using three external datasets. Receiver operating characteristic analyses were performed to evaluate the diagnostic performance to determine active TB by DL model and radiologists. Sensitivities and specificities for determining active TB were evaluated for both the DL model and radiologists. Results: The performance of the DL model showed area under the curve (AUC) values of 0.980 in internal validation, and 0.815 and 0.887 in external validation. The AUC values for the DL model, thoracic radiologist, and general radiologist, evaluated using one of the external validation datasets, were 0.815, 0.871, and 0.811, respectively. Conclusion: This DL-based algorithm showed potential as an effective diagnostic tool to identify TB activity, and could be useful for the follow-up of patients with inactive TB in high TB burden countries.