http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
이다빈,신태주,유필진,오경화,박주현 한국공업화학회 2018 Journal of Industrial and Engineering Chemistry Vol.63 No.-
We prepared conjugated polymer nano-ellipsoids (CPNs) via emulsification of chloroform phase using the octanoic acid (OA) in dimethyl formamide (DMF), followed by the removal of chloroform by heating. The resulting CPN DMF solutions were thoroughly mixed with polyurethane (PU) DMF solutions to form composite films upon solvent removal, with uniformly distributed CPNs due to hydrogen bonds between PU matrix and the CPNs. Superior photothermal and antibacterial properties of the PU:CPN composite films were observed, presenting the usefulness of CPNs as an efficient light harvester and thermal storage material, and the OA as an antibacterial material for multifunctional fiber applications.
Antioxidant and Cytoprotective Effects of Lotus (Nelumbo nucifera) Leaves Phenolic Fraction
이다빈,김도형,제재영 한국식품영양과학회 2015 Preventive Nutrition and Food Science Vol.20 No.1
Phenolic rich ethyl acetate fraction (EAF) from lotus leaves was prepared and its bioactive components, antioxidant and cytoprotective effects were investigated. EAF showed high total phenolic content and flavonoid content and contained rutin (11,331.3±4.5 mg/100 g EAF), catechin (10,853.8±5.8 mg/100 g EAF), sinapic acid (1,961.3±5.6 mg/100 g EAF), chlorogenic acid (631.9±2.3 mg/100 g EAF), syringic acid (512.3±2.5 mg/100 g EAF), and quercetin (415.0±2.1 mg/100 g EAF). EAF exerted the IC50 of 4.46 mg/mL and 5.35 mg/mL toward DPPH and ABTS cation radicals, respectively, and showed strong reducing power, which was better than that of ascorbic acid, a positive control. Additionally, EAF protected hydroxyl radical-induced DNA damage indicated by the conversion of supercoiled pBR322 plasmid DNA to the open circular form and inhibited lipid peroxidation of polyunsaturated fatty acid in a linoleic acid emulsion. In cultured hepatocytes, EAF exerted a cytoprotective effect against oxidative stress by inhibiting intracellular reactive oxygen species formation and membrane lipid peroxidation. In addition, depletion of glutathione under oxidative stress was remarkably restored by treatment with EAF. The results suggest that EAF have great potential to be used against oxidative stress-induced health conditions.
랜드마크 유전자 기반 목표 유전자 발현량 예측을 위한 특징 추출 심층 신경망
이다빈,황규백 한국정보과학회 2020 정보과학회 컴퓨팅의 실제 논문지 Vol.26 No.8
Gene expression profiling is useful for disease studies. Researchers of the Library of Integrated Network-Based Cellular Signatures Program developed an efficient profiling method in which the expression level of only a subset of the human genes (landmark genes), comprising approximately 80% of the entire gene expression information, is measured, and subsequently used for predicting the expression level of the other genes (target genes). In this study, we propose a method to extract non-linear features of the landmark genes using autoencoders and then predict the expression level of the target genes using the extracted features. In the experiments on 111,009 gene-expression profiles, comprising 943 landmark genes and 9,520 target genes, our method reduced the prediction error for approximately 95% of the target genes compared with a previous deep neural network model. The average proportion of error reduction was approximately 7%. This result suggests that the non-linear feature extraction can improve the accuracy of target-gene expression prediction from the landmark genes. 유전자 발현 프로파일링은 질병 연구에 유용하다. Library of Integrated Network-Based Cellular Signatures 프로그램 연구진은 전체 인간 유전자 발현 정보의 약 80%를 포함하는 소수의 유전자(랜드마크 유전자) 발현량만을 측정한 뒤 다른 유전자(목표 유전자)의 발현량을 예측하는 효율적인 프로파일링 기법을 개발했다. 본 논문에서는 오토인코더로 랜드마크 유전자의 비선형 특징을 추출한 후 이에 기반하여 목표 유전자의 발현량을 예측하는 방법을 제안한다. 이 방법은 943개의 랜드마크 유전자와 9,520개의 목표 유전자로 구성된 111,009개의 유전자 발현 프로파일에 대한 실험에서 기존의 심층 신경망과 비교했을 때 약 95%의 목표 유전자에 대해 예측 오류를 감소시켰으며, 감소의 폭은 평균적으로 약 7%였다. 이러한 결과는 오토인코더 기반의 비선형 특징 추출이 랜드마크 유전자로부터 목표 유전자의 발현량을 예측하는 데 기여할 수 있음을 시사한다.
딥 러닝 기반 road segmentation module의 일반화된 성능 확보를 위한 학습 데이터의 조건 분석
이다빈,김승현,나문수,이홍준,김제연,김회율 대한전자공학회 2019 대한전자공학회 학술대회 Vol.2019 No.6
Road segmentation 문제를 해결하기 위해 딥 러닝이 각광받고 있다. 딥 러닝 모델의 학습을 위해서는 도로를 주행하며 촬영된 도로 데이터 셋이 필요한데 일반적으로 한 대의 차량에서 차량 전방에 설치된 카메라로 데이터를 취득한다. 일부 공용 데이터 셋들로 같은 방법으로 데이터를 취득하나 이러한 데이터 셋을 사용해 학습한 모델은 하나의 시점에서 바라본 도로 영역에 대한 학습만이 이루어지기 때문에 다른 시점으로 도로 영역을 촬영한 데이터 셋에 대하여 road segmentation 성능이 떨어지는 문젝 발생한다. 본 논문은 동일한 시점의 데이터 셋으로 학습시킨 모델과 다양한 시점의 데이터 셋으로 학습시킨 모델을 비교하여 다양한 시점으로 촬영된 학습 데이터 셋을 사용한 모델이 하나의 시점에서만 촬영된 학습 데이터 셋을 사용한 모델보다 일반화된 성능을 보임을 확인하고 그 원인을 분석한다.
이다빈,이용주,강민숙,김행란,장경아 한국식품조리과학회 2022 한국식품조리과학회지 Vol.38 No.6
Purpose: The purpose of this study is to analyze the resistant starch(RS) content of legumes and to develop soymilk with increased resistant starch. Methods: To estimate the functional effect of leguemes containing RS, the starch content(digestable starch, total starch) a hydrolysis index, and a glycemic index are established for boiled legumes and manufactured soymilk, since the RS content changes depending on the processing method. Results: From among the boiled legumes, kidney beans showed the highest RS content, followed by mungbeans, chickpeas, lentils, and soybeans. The hydrolysis index and the glycemic index of D-glucose were highest in soybeans and lowest in kidney beans. Also, RS in manufactured soymilk was highest in kidney beans, and kidney bean soy milk(20-mesh sieve) showed an especially high content of RS compared to cotton-filtered and commercial soybean soy milk(the control). Kidney bean soy milk showed higher carbohydrate and protein content but lower fat content than the control. Conclusion: We confirm that RS content changes during cooking and filtration. The results of this study, can be used as basic data to develop functional soymilk with high RS.
In Vitro Digestion Model을 활용한 두류 소화 단계별 영양성분 변화 분석
이다빈,장경아,황인선,강민숙,서미경,김행란,유선미 한국식품영양학회 2023 韓國食品營養學會誌 Vol.36 No.5
Changes in contents of free sugars, amino acids, and fatty acids of legumes were analyzed for each phase of in vitro digestion. In addition, contents of resistant starch in raw and digested pulses were compared. Soybeans, kidney beans, cowpeas, and chickpeas were analyzed. An in vitro digestion model was used to analyze contents of nutrients using LC-MS and GC-MS. Stachyose in kidneybean, cowpea, and chickpea increased as the digestion phase progressed. In four types of legumes, raffinose slightly decreased or showed no significant difference between the Oral phase and the BBMV phase. Content of glucose, a monosaccharide, increased during the BBMV phase. During the digestion phase, levels of free amino acids and free fatty acids also increased. Content of resistant starch was reduced compared to that in the raw material. It was 0.01g/100 g food in soybean, 1.06 g/100 g food in red kidney bean, 0.77g/ 100g food in cowpea, and 0.76 g/100 g food in chickpea. It was confirmed that nutrients in the in vitro digestion model were liberated at each digestion phase with changes in the content of resistant starch. These results are expected to be used as fundamental data for obtaining bioavailability of nutrients.