단일 스텐드 냉간 압연 공정에 대한 자동조정 퍼지 논리 장력제어기의 설계 ( Design of an Auto-Tuning Fuzzy Logic Tension Controller for a Single-Stand Cold Rolling Mill Process )

정윤기,안현식,김도현 대한전자공학회 1996 대한전자공학회 학술대회 Vol.1996 No.7

CD34 Monoclonal Antibody-Immobilized Electrospun Polyurethane for the Endothelialization of Vascular Grafts

정윤기,황인규,박기동,이찬우 한국고분자학회 2010 Macromolecular Research Vol.18 No.9

Targeting endothelial progenitor cells (EPC) for in vivo endothelialization is an emerging and promising approach for the development of cardiovascular medical devices. This study examined the efficacy of capturing CD34 positive EPC onto polyurethane (PU) immobilized with CD34 monoclonal antibodies (mAbs) (a biomecial polymer for cardiovascular devices). Electrospun PU matrices were fabricated and heparin was immobilized along with CD34 mAb. The modified PU surfaces at each step were characterized by contact angle measurements,atomic force microscopy (AFM), and X-ray photoelectron spectroscopy (XPS). XPS showed that each surface was modified, as expected in terms of the chemical composition. The amine-terminated poly(ethylene glycol)(PEG)-PU surface was considerably more hydrophilic than the PU surface. In addition, its surface roughness was similar to the PU surface, indicating that PEG was sufficiently and evenly grafted onto the PU surface. The CD34 mAb-immobilized PEG-PU surface was less hydrophilic than PEG-PU and extremely rough as compared to the other two surfaces. These results demonstrate that relatively large CD34 mAbs were immobilized on the PU surface. The surface density of the immobilized CD34 mAb, which was quantified using an enzymelinked immune-sorbent assay (ELISA), was increased to ~40 ng/cm2 by varying the feed amount up to ~200 ng/cm2 and co-immobilizing with heparin. These results suggest that the co-immobilization with heparin can provide two benefits: inhibiting initial occlusion and improving the surface density of CD34 mAb. The in vitro cell study also demonstrated that the CD34 mAb-immobilized PU surface was favorable for cell attachment and proliferation. Therefore, in this study, a novel approach was developed to achieve endothelialization for cardiovascular applications by immobilizing CD34 onto PU, and the synergistic effects of co-immobilization with heparin on the bioactivity of the PU surfaces was demonstrated.

Universal Coating of Robust Cell-Derived Extracellular Matrices for Implantable Materials

정윤기 한국고분자학회 2017 한국고분자학회 학술대회 연구논문 초록집 Vol.42 No.2

자연어를 활용한 SQL문 생성을 위한 합성곱 신경망 기반 칼럼 예측 모델

정윤기,김동민,이종욱 한국정보과학회 2019 정보과학회논문지 Vol.46 No.2

To retrieve massive data using relational database management system (RDBMS), it is important to understanding of table schemas and SQL grammar. To address this issue, many studies have recently been carried out to generate an SQL query from a natural language question. However, the existing works suffer mostly from predicting columns at where clause and the accuracy is greatly reduced when there are multiple columns to be predicted. In this paper, we propose a convolutional neural network model with column attention mechanism that effectively extracts the latent representation of input question which helps column prediction of the model. The experiment shows that our model outperforms the accuracy of the existing model (SQLNet) by 6%. 관계형 데이터베이스 시스템을 이용하여 대규모의 데이터를 검색하기 위해서는 테이블 스키마 및 SQL문을 이해해야 하는 필요성이 있다. 이를 해결하기 위해 자연어가 입력으로 주어질 때, 이에 대응하는 SQL문을 생성하는 연구가 최근 진행되고 있다. 기존 연구에서 가장 어려운 부분은 SQL문의 조건에 해당되는 칼럼을 효과적으로 예측하는 부분이며, 예측해야 하는 칼럼의 개수가 여러 개일 때 정확도가 크게 떨어지는 문제점이 있다. 본 논문에서는 칼럼 어텐션 메카니즘을 이용하여, 자연어 데이터의 숨겨진 표현을 효과적으로 추출하는 합성곱 신경망 모델을 제안한다. 본 연구의 제안 방법은 기존 방법 대비 약 6% 이상 정확도가 향상되는 것을 확인할 수 있었다.

운동 형태에 따른 에너지 효율과 렙틴 호르몬

정윤기,김현 한국스포츠리서치 2004 한국 스포츠 리서치 Vol.15 No.1

Variations of Resting-State EEG-Based Functional Networks in Brain Maturation From Early Childhood to Adolescence

정윤기,전용훈,김려경,조안나,김헌민,황희,최지은,김기중 대한신경과학회 2022 Journal of Clinical Neurology Vol.18 No.5

Background and Purpose Alterations in human brain functional networks with maturation have been explored extensively in numerous electroencephalography (EEG) and functional magnetic resonance imaging studies. It is known that the age-related changes in the functional networks occurring prior to adulthood deviate from ordinary trajectories of networkbased brain maturation across the adult lifespan. Methods This study investigated the longitudinal evolution of resting-state EEG-based functional networks from early childhood to adolescence among 212 pediatric patients (age 12.2± 3.5 years, range 4.4–17.9) in 6 frequency bands using 8 types of functional connectivity measures in the amplitude, frequency, and phase domains. Results Electrophysiological aspects of network-based pediatric brain maturation were characterized by increases in both functional segregation and integration up to middle adolescence. EEG oscillations in the upper alpha band reflected the age-related increases in mean node strengths and mean clustering coefficients and a decrease in the characteristic path lengths better than did those in the other frequency bands, especially for the phase-domain functional connectivity. The frequency-band-specific age-related changes in the global network metrics were influenced more by volume-conduction effects than by the domain specificity of the functional connectivity measures. Conclusions We believe that this is the first study to reveal EEG-based functional network properties during preadult brain maturation based on various functional connectivity measures. The findings potentially have clinical applications in the diagnosis and treatment of age-related brain disorders

Acellular surface-modified scaffolds for delivery of nephron progenitor cell for renal tissue regeneration

정윤기,김윤아,( Md Lemon Hasan ),( So Young Chun ),( Seung Kwon You ),( Tae Gyun Kwon ),한동근,( Won-gun Koh ) 한국공업화학회 2020 한국공업화학회 연구논문 초록집 Vol.2020 No.-

In recent year, renal tissue regeneration with the help of extracellular matrix (ECM) derived scaffold is expected to be a promising strategy. Extracellular matrix (ECM) derived scaffold owing to its highly biocompatiblity and ability to guid renel tissue regeneration. In our previous study, magnesium hydroxide nanoparticles incorporated three dimensional porous PLGA scaffold was directly coated with fibroblast derived extracellular matrix. Inspired from previous study, here we have prepared nephron progenitor cell containing ECM surface- modified scaffolds for renal tissue regeneration. In in vitro cell studies with NPCs, modified scaffold showed improved cell viability and significantly increased growth factor release in respective time. An 8-week post implantation study conducted on kidney model demonstrated that modified scaffold reduced the expression of fibrogenesis inducing cytokine marker and accelerated the renal tissue regeneration with a low inflammatory response.

신경망 및 비신경망 오토인코더 기반 추천 모델의 성능 비교 및 분석

정윤기,이종욱 한국정보과학회 2020 정보과학회논문지 Vol.47 No.11

다양한 분야에 심층 신경망이 도입되어 획기적인 성능 개선을 보이고 있으나, 최근 심층 신경망 기반 추천 모델의 성능 개선이 크게 보이지 않는다는 주장이 나오고 있다. 이와 같은 문제는 추천 연구에 통용되는 실험 환경의 부재와 제안 모델 성능에 대한 엄밀한 분석 부재에 기인한다. 본 논문에서는 1) 추천 모델의 공정한 비교를 위한 실험 프로토콜을 구성하고, 2) 추천 모델의 한 축인 오토인코더 기반 추천 모델에 대해서 실험적 검증을 수행하며, 3) 사용자와 항목 인기도를 기준으로 여러 개의 세부 그룹으로 나누어 실험 결과를 분석한다. 실험 결과, 모든 데이터셋에서 신경망 기반 모델의 추천 성능이 비신경망 대비 일관적인 성능 개선을 보이지 않았으며, 신경망 모델 내에서도 주된 정확도 개선을 확인할 수 없었다. 한편, 세부 그룹별 성능 평가 결과에서는 인기 항목에선 비신경망 모델의, 비인기 항목에선 신경망 모델의 정확도가 높음이 확인하였고, 이를 통해 신경망 모델의 복잡성이 비인기 항목에 대한 추천에 도움이 될 수 있다고 판단된다. While deep neural networks have been bringing advances in many domains, recent studies have shown that the performance gain from deep neural networks is not as extensive as reported, compared to the higher computational complexity they require. This phenomenon is caused by the lack of shared experimental settings and strict analysis of proposed methods. In this paper, 1) we build experimental settings for fair comparison between the different recommenders, 2) provide empirical studies on the performance of the autoencoder-based recommender, which is one of the main families in the literature, and 3) analyze the performance of a model according to user and item popularity. With extensive experiments, we found that there was no consistent improvement between the neural and the non-neural models in every dataset and there is no evidence that the non-neural models have been improving over time. Also, the non-neural models achieved better performance on popular item accuracy, while the neural models relatively perform better on less popular items.

Surface Modification of Metallic Stent using Estrogen for the Prevention of Restenosis

정윤기,박기동,김현일,김성수 한국고분자학회 2002 한국고분자학회 학술대회 연구논문 초록집 Vol.2002 No.4

조직재생을 위한 고분자 지지체의 최근 연구개발 동향

정윤기,박기동,박귀덕,한동근,Joung, Yoon-Ki,Park, Ki-Dong,Park, Kwi-Deok,Han, Dong-Keun 대한의용생체공학회 2008 의공학회지 Vol.29 No.4

In tissue engineering, scaffolds play an important role in the growth of cells to 3-D organs or tissues. For the success of tissue engineering, they should be mimicked to meet the requirements of natural extracellular matrix (ECM) in the body, such as mechanical properties, adhesiveness, porosity, biodegradability, and growth factor release, etc. Contrary to other materials, polymeric materials are adequate to engineer scaffolds for tissue engineering because controlling the structure and the ratio of components and designing various shapes and size are possible. In this review, the importance, major characteristics, processes, and recent examples of polymeric scaffolds for tissue engineering applications are discussed.