• Geomechanics and Engineering
• /
• 제37권5호
• /
• pp.467-474
• /
• 2024

Monitoring and managing the condition of underground utilities is crucial for ground stability. This study aims to determine whether images obtained using ground penetrating radar (GPR) accurately reflect the characteristics of buried pipelines through image analysis. The investigation focuses on pipelines made from different materials, namely concrete and steel, with concrete pipes tested under various diameters to assess detectability under differing conditions. A total of 400 images are acquired at locations with pipelines, and for comparison, an additional 100 data points are collected from areas without pipelines. The study employs GPR at frequencies of 200 MHz and 600 MHz, and image analysis is performed using machine learning-based convolutional neural network (CNN) techniques. The analysis results demonstrate high classification reliability based on the training data, especially in distinguishing between pipes of the same material but of different diameters. The findings suggest that the integration of GPR and CNN algorithms can offer satisfactory performance in exploring the ground's interior characteristics.

• Nuclear Engineering and Technology
• /
• 제56권2호
• /
• pp.745-752
• /
• 2024

A method for gamma/neutron event classification based on frequency-domain analysis for mixed radiation environments is proposed. In contrast to the traditional charge comparison method for pulse-shape discrimination, which requires baseline removal and pulse alignment, our method does not need any preprocessing of the digitized data, apart from removing saturated traces in sporadic pile-up scenarios. It also features the identification of neutron events in the detector's full energy range with a single device, from thermal neutrons to fast neutrons, including low-energy pulses, and still provides a superior figure-of-merit for classification. The proposed frequency-domain analysis consists of computing the fast Fourier transform of a triggered trace and integrating it through a simplified version of the transform magnitude components that distinguish the neutron features from those of the gamma photons. Owing to this simplification, the proposed method may be easily ported to a real-time embedded deployment based on Field-Programmable Gate Arrays or Digital Signal Processors. We target an off-the-shelf detector based on a small CLYC (Cs2LiYCl6:Ce) crystal coupled to a silicon photomultiplier with an integrated bias and preamplifier, aiming at lightweight embedded mixed radiation monitors and dosimeter applications.

• Earthquakes and Structures
• /
• 제27권2호
• /
• pp.83-95
• /
• 2024

The prediction of seismic behavior of the existing building stock is one of the most impactful and complex problems faced by countries with frequent and intense seismic activities. Human lives can be threatened or lost, the economic life is disrupted and large amounts of monetary reparations can be potentially required. However, authorities at a regional or national level have limited resources at their disposal in order to allocate to preventative measures. Thus, in order to do so, it is essential for them to be able to rank a given population of structures according to their expected degree of damage in an earthquake. In this paper, the authors present a ranking approach, based on Machine Learning (ML) algorithms for pairwise comparisons, coupled with ad hoc ranking rules. The case study employed data from 404 reinforced concrete structures with various degrees of damage from the Athens 1999 earthquake. The two main components of our experiments pertain to the performance of the ML models and the success of the overall ranking process. The former was evaluated using the well-known respective metrics of Precision, Recall, F1-score, Accuracy and Area Under Curve (AUC). The performance of the overall ranking was evaluated using Kendall's tau distance and by viewing the problem as a classification into bins. The obtained results were promising, and were shown to outperform currently employed engineering practices. This demonstrated the capabilities and potential of these models in identifying the most vulnerable structures and, thus, mitigating the effects of earthquakes on society.

• 한국컴퓨터정보학회:학술대회논문집
• /
• 한국컴퓨터정보학회 2021년도 제64차 하계학술대회논문집 29권2호
• /
• pp.305-307
• /
• 2021

Alzheimer's Disease (AD) is a cognitive disorder characterized by memory impairment that can be assessed at early stages based on administering clinical tests. However, the AD pathophysiological mechanism is still poorly understood due to the difficulty of distinguishing different levels of AD severity, even using a variety of brain modalities. Therefore, in this study, we present a hybrid EEG-fNIRS modalities to compensate for each other's weaknesses with the help of Machine Learning (ML) techniques for classifying four subject groups, including healthy controls (HC) and three distinguishable groups of AD levels. A concurrent EEF-fNIRS setup was used to record the data from 41 subjects during Oddball and 1-back tasks. We employed both a traditional neural network (NN) and a CNN-LSTM hybrid model for fNIRS and EEG, respectively. The final prediction was then obtained by using majority voting of those models. Classification results indicated that the hybrid EEG-fNIRS feature set achieved a higher accuracy (71.4%) by combining their complementary properties, compared to using EEG (67.9%) or fNIRS alone (68.9%). These findings demonstrate the potential of an EEG-fNIRS hybridization technique coupled with ML-based approaches for further AD studies.

• 정보처리학회논문지:소프트웨어 및 데이터공학
• /
• 제6권11호
• /
• pp.537-542
• /
• 2017

공격적인 성향의 운전은 자동차 사고의 주요한 원인이 된다. 기존 연구에서는 공격적 성향의 운전을 검출하기 위해, 주로 청년을 대상으로 연구가 이뤄졌으며 기계학습의 순수한 Clustering 또는 Classification 기법을 통해 이뤄졌다. 그러나 노인들은 취약한 신체적 조건에 의해 젊은 운전자와는 다른 운전 강도를 가지고 있어 기존의 방식으로는 검출이 불가능 하며, 데이터를 보정하는 등의 새로운 방법이 필요하다. 그리하여, 본 연구에서는 기존의 클러스터링 기법(K-means, Expectation - maximization algorithm)에, 새롭게 제안하는 ECA(Enhanced Clustering method for Acceleration data)기법을 추가하여, 주행 차량에 위치한 스마트폰으로부터 수집된 가속도 데이터를 분석하고 공격적인 운전 형태를 검출해 낸다. ECA는 모든 피험자의 데이터에서 K-means와 EM을 통해 검출된 군집군의 데이터 중 높은 강도의 데이터를 선별하여, 특징을 스케일링한 값을 통해 모델링한다. 본 방식을 통해 기존의 연구의 순수한 클러스터링 방식과는 달리, 모든 청장년 및 노인 실험 참가자 개인들의 공격적인 운전 데이터가 검출되었으며, 클러스터링 기법간의 비교를 통해 K-means 기법이 보다 높은 검출 효율을 갖고 있음을 확인했다. 또한, K-means 방식을 검출한 공격적인 운전 데이터에서는 젊은 운전자가 노인운전자에 비해 1.29배의 높은 운전 강도를 가지고 있음을 발견했다. 이와 같이 본 연구에서 제안된 방식은 낮은 운전 강도를 갖고 있는 노인의 데이터에서 공격적인 운전을 검출 가능하게 되었으며, 특히. 제안된 방법은 노인 운전자를 위한 맞춤형 안전운전 시스템을 구축이 가능하며, 추후 다양한 연구을 통해 이상 운전 상태를 검출하고 조기 경보하는데 활용이 가능할 것이다.

• Genomics & Informatics
• /
• 제14권2호
• /
• pp.46-52
• /
• 2016

Diverse somatic mutations have been reported to serve as cancer drivers. Recently, it has also been reported that epigenetic regulation is closely related to cancer development. However, the effect of epigenetic changes on cancer is still elusive. In this study, we analyzed DNA methylation data on colon cancer taken from The Caner Genome Atlas. We found that several promoters were significantly hypermethylated in colon cancer patients. Through clustering analysis of differentially methylated DNA regions, we were able to define subgroups of patients and observed clinical features associated with each subgroup. In addition, we analyzed the functional ontology of aberrantly methylated genes and identified the G-protein-coupled receptor signaling pathway as one of the major pathways affected epigenetically. In conclusion, our analysis shows the possibility of characterizing the clinical features of colon cancer subgroups based on DNA methylation patterns and provides lists of important genes and pathways possibly involved in colon cancer development.

• Ecology and Resilient Infrastructure
• /
• 제5권2호
• /
• pp.59-71
• /
• 2018

Lake Enriquillo, the largest lake in the Dominican Republic, recently has undergone unusual water area changes since 2001 thus it has been affected seriously by local community's livelihood. Earthquakes and seismic activities of Hispaniola plate tectonic coupled with human activities and climate change are addressed as factors causing the increasing. Thus, a thorough study on relationship between lake area changing, and those factors is needed urgently. To do so, this study applied MESMA on MODIS data to extract water area of Lake Enriquillo during 2001 and 2012 bimonthly, with six issues 12-year. MODIS provides high temporal resolution, and its coarse spatial resolution is compensated by MESMA fraction map. The increase in water area was $142.2km^2$, and the maximum lake area was $338.0km^2$ (in 2012). Water areas extracted by two Landsat scenes at two different times with three image classification approaches (ISODATA, MNDWI, and TCW) were used to assess accuracy of MODIS and MESMA results; it indicated that MESMA water areas are same as ISODATA's, less than 0.4%, while the highest difference is between MESMA and TCW, 2.4%. A number of previously formulated hypotheses of lake area change were investigated based on the outcomes of the present study, though none of them could fully explain the changes.

• 한국수자원학회:학술대회논문집
• /
• 한국수자원학회 2021년도 학술발표회
• /
• pp.6-7
• /
• 2021

하천에서의 식생 활착은 지형, 생태, 수리학 등의 학문 분야 뿐만 아니라 하천 관리 실무에서도 중요한 이슈 중에 하나로서 하천 식생 문제는 홍수 관리와 생태계 보전이라는 상반되는 가치의 조화에 직결된다. 국내에서는 2000년대 이후 댐 하류 조절하천, 부영화된 소규모 지류하천, 4대강 사업 대상지 고수부지 등 다양한 조건에서 하천 식생 활착과 육역화 문제가 지속적으로 제기되어 왔다. 이러한 배경에서 본 연구에서는 하천 내의 식생 분포를 원격탐사 자료를 기반으로 분류하는 기법을 제안하고 이를 내성천에 적용한 결과를 제시하였다. 내성천은 2014년부터 최근까지 지속적으로 식생 활착이 발생하여 하천 경관이 변화한 대표적인 사례 하천이다. 원격탐사 자료는 유럽항공우주국(ESA)에서 운영 중이며, Google Earth Engine에서 제공하는 Sentinel 1, 2 위성 영상을 사용하였다. 지상 참값(ground truth)으로는 수역, 사주, 초본, 목본 등을 포함한 8가지 유형으로 구분되어 있는 2016년 내성천 지표 피복 자료를 사용하였다. 분류를 위한 방법은 머신러닝 알고리듬의 하나인 랜덤 포레스트 분류 기법을 사용하였으며, 미리 선정된 10개 폴리곤 영역으로부터 1,000개의 표본을 추출하여 1/2씩 나누어 훈련 및 검증 자료로 사용하였다. 검증 자료 기반의 정확도는 82~85 %로 나타났다. 훈련을 통해 수립한 모형을 2016~2020년 자료에도 적용하여 연도에 따른 식생역의 변화 과정을 제시하였다. 본 논문의 기술적 한계와 개선 방안을 고찰하였다. 이 기법은 정량적인 식생 분포를 제공함으로써 하천에서의 홍수위 계산, 식생-수리모델링 등의 기술 분야 뿐만 아니라 간벌이나 하천 식생 회춘 유도(rejuvenation)과 같은 식생의 실무적 관리 측면에서도 활용도가 클 것으로 판단된다.

• Journal of Microbiology and Biotechnology
• /
• 제21권4호
• /
• pp.421-429
• /
• 2011

For precise identification of a Lactobacillus K1 isolate, LC-MS/MS analysis of the putative surface layer protein was performed. The results obtained from LTQ-FT-ICR mass spectrometry confirmed that the analyzed protein spot is the surface layer protein originating from Lb. helveticus species. Moreover, the identified protein has the highest similarity with the surface layer protein from Lb. helveticus R0052. To evaluate the proteomic study, multilocus sequence analysis of selected housekeeping gene sequences was performed. Combination of 16S rRNA sequencing with partial sequences for the genes encoding the RNA polymerase alpha subunit (rpoA), phenylalanyl-tRNA synthase alpha subunit (pheS), translational elongation factor Tu (tuf), and Hsp60 chaperonins (groEL) also allowed to classify the analyzed isolate as Lb. helveticus. Further classification at the strain level was achieved by sequencing of the slp gene. This gene showed 99.8% identity with the corresponding slp gene of Lb. helveticus R0052, which is in good agreement with data obtained by nano-HPLC coupled to an LTQ-FT-ICR mass spectrometer. Finally, LC-MS/MS analysis of surface layer proteins extracted from three other Lactobacillus strains proved that the proposed method is the appropriate molecular tool for the identification of S-layer-possessing lactobacilli at the species and even strain levels.

• 전자공학회논문지C
• /
• 제36C권12호
• /
• pp.90-97
• /
• 1999

최근에 컴퓨터를 이용한 영상처리기술 및 고속통신망의 발달과 더불어 하드웨어의 고성능화로 의학분야에서 발생되는 영상들에 대해 분석 및 처리를 자동화하려는 많은 연구가 진행되고 있다. 본 논문에서는 말초혈액영상에서 혈구세포들을 자동으로 분석, 분류 및 카운트하기 위해 다층신경망에 기반한 시스템을 설계 및 구현하였다. 이를 위해 먼저 CDD 카메라가 부착된 현미경으로부터 영상을 입력받아 적혈구와 백혈구 분류를 위한 다양한 특징추출 알고리즘을 적용하였다. 또한, PCA를 적용해 다차원의 특징을 저차원으로 줄여 분류기의 훈련과 인식 시간을 단축시킴으로서 보다 효율적인 분류기 시스템을 구축하였다. 따라서 , 본 논문에서는 제안된 시스템이 실제 임상 병리진단 가이드 시스템에 적용 가능함을 보일 수 있었다.