• Proceedings of the Korea Information Processing Society Conference
• /
• 2021.11a
• /
• pp.577-579
• /
• 2021

의료영상기반의 인공지능 연구는 질환의 조기진단 및 예측 분야에 눈부신 기술발전이 되어왔다. 근감소증 질환은 다양한 기저질환을 기반으로 발생하며, 특히 60대 이상은 30%의 유병율을 갖는다. 해당 질환은 임상적인 진단 방법의 발달과 임상 결과가 알려지면서 관심이 증가하고 있다. 최근 근감소증 진단방법 중의 하나로 CT 또는 MR 의료영상을 통한 진단방법이 제시되었다. 본 논문에서는 인공지능을 기반으로 하여, 근감소증을 진단하기 위해 척추부위 중 Lumbar 3 영역의 근육, 지방 영역의 영상분할 모델을 제시하고자 한다. 이를 위해 인공지능 영상분할 모델을 개발하는 과정과 그 근육과 지방의 영상분할 결과를 보인다. 본 논문에서 제시한 영상분할모델을 통해 근감소증을 빠르게 진단할 수 있을 것으로 기대한다.

• Proceedings of the Membrane Society of Korea Conference
• /
• 1994.04a
• /
• pp.26-27
• /
• 1994

효소는 생체내의 합성, 분해, 산화, 환원 등 복잡한 화학반응이 상온, 상압, 중성부근에서 효율적으로 진행되게 하는 단백질이 주성분인 유기촉매이다. 현재 알려져 있는 효소의 종류는 수백만종 이상으로 추정되며 그 중 100여종 이상은 순수한 결정상태이며 약 600종 정도는 어느 정도 순수하게 징제되고 있다. 이들 효소의 분자량은 Ribonuclease의 12,700에서 부터 L-Glutamate dehydrogenase나 Carboxylase의 1,000,000 이상으로 광범위하다.

• Proceedings of the Korean Society for Agricultural Machinery Conference
• /
• 2017.04a
• /
• pp.166-166
• /
• 2017

모돈의 건강 상태를 정량 지수화 하기 위한 연구를 수행 중이다. 지제이상, 섭식 불량, 수면 패턴 등의 운동 특성 분석을 위하여 복수의 초음파 센서를 이용하였다. 시계열 계측 신호를 분석하여 정량 지수화를 수행하는 과정에서 주파수 도메인 분석을 시도하였다. 이 과정에서 주파수 도메인의 분해능에 따른 편차 극복을 위한 비선형 모델링을 수행하였다. 또한 인접한 시계열 데이터 구간 간의 상관성 분석이 가능하면 대용량 데이터의 실시간 처리로 인한 지연 시간 극복 및 기대되는 예후에 대한 조기 진단이 가능할 것이다. 본 연구에서는 구글에서 제공하는 Tensorflow와 NVIDIA에서 제공하는 CUDA 엔진을 동시 적용한 심층 학습 시스템을 이용하였다. 전 처리를 위하여 주파수 분해능 (2분, 3분, 5분, 7분, 11분, 13분, 17분, 19분)에 따른 데이터 집합을 1단계로 두고, 상위 10 순위 안에 드는 파워 스펙트럼 밀도의 크기를 2단계로 하여, 총 2~10개의 입력 노드를 순차적으로 선정하였고, 동일한 방식으로 인접한 시계열의 파워 스펙터럼 밀도를 순위를 변화시켜 지정하였다. 대표적인 심층학습 모델인 Softmax regression with a multilayer convolutional network를 이용하여 Recursive feature selection 경우의 수를 $8{\times}9{\times}9$로 총 648 가지 선정하고, Epoch는 10,000회로 지정하였다. Calibration 모델링의 경우 Cost function이 10% 이하인 경우 해당 경우의 학습을 중단하였으며, 모델 간 상호 교차 검증을 수행하기 위하여 $_8C_2{\times}_8C_2{\times}_8C_2$ 경우의 수에 대한 Verification test를 수행하였다. Calibration 과정 상 모든 경우에 대하여 10% 이하의 Cost function 값을 보였으나, 검증 테스트 과정에서 모든 경우에 대하여 $r^2$ < 0.5 인 결정 계수 값이 나타났다. 단적으로 심층학습 모델의 과도한 적합(Over fitting) 방식의 한계를 보인 것이라 판단할 수 있다. 적합한 Feature selection 및 심층 학습 모델에 대한 지속적이고 추가적인 고려를 통해 과도적합을 해소함과 동시에 실효적이고 활용 가능한 Classification을 위한 입, 출력 노드 단의 전후 Indexing, Quantization에 대한 고려가 필요할 것이다. 이를 통해 모돈 생체 정보 정량화를 위한 지능형 현장 진단 기술 연구를 지속할 것이다.

• Journal of the Korean Society of Food Science and Nutrition
• /
• v.23 no.5
• /
• pp.731-735
• /
• 1994

The aims of this study are conducted to investigate the effect of pretreatment with Achatina fucica extract (AFE) in streptozotocin (STZ)-induced diabetic rats. Serum glucose, total lipid and triglyceride levels of administration group of AE in diabetic rats induced by STZ were significant high levels than that of control group. However, glucose, triglyceride and lipid levels were significant lower level in group of pretreatment with AFE and Atheroslerotic index decreased in the group of treatment of AFE when those levels compared with that of STZ-treated group. Serum lipase activity was inhibited in the control group STZ induced diabetic rats, in contrast lipase activity increased in the group of AFE administration. These results suggested that AFE may use to prevent the diabetes mellitus induced by STZ.

• Journal of the Korean Institute of Telematics and Electronics T
• /
• v.36T no.4
• /
• pp.1-8
• /
• 1999

This paper presents a new algorithm for surface segmentation and feature description. In the first stage of proposed algorithm, the signature of an edge image of object is extracted. The signature technique represents a surface using the distance from the mass center to the boundary of the image as a function of angle rotating counterclockwise. If there exists a range in the angle axis where more than two signatures form a closed curve, we can conclude there is a surface inside the range. Using this feature of the signature, surface can be segmented. The surface features such as number of vertices, number of edges, convex and type of surface can also be extracted from segmented surfaces. This algorithm has distinguished advantages; it can easily recover the lost part in the edge image using the curve fitting method; it extracts surface features correctly regardless of the rotation of the surface in 3-D space.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.35 no.2
• /
• pp.319-326
• /
• 2015

Rainfall data is necessary component for water resources design and flood warning system. Most analysis are used long-term hourly data of surface synoptic stations from the Meteorological Administration, Ministry of land, Infrastructure and Transport and others. However, It will be used minutely data of more high density automatic weather stations than surface synoptic stations expecting to increase the frequency of heavy precipitation. But minutely data has a problem about quality of rainfall data by auto observation. This study analyzed about quality control method using automatic weather station's minutely rainfall data of meteorological administration. It was performed assessment of the quality control that was classified quality control of miss Data, outlier data and rainfall interpolation. This method will be utilized when hydrological analysis uses minute rainfall data.

• The KIPS Transactions:PartD
• /
• v.13D no.7 s.110
• /
• pp.867-878
• /
• 2006

In the past few years, the research of sensor networks is forced dramatically. Specially, while the research for maintaining the power of a sensor is focused, we are also concerned nth query processing related with the optimization of multiple continuous queries for decreasing in unnecessary energy consumption of sensor networks. We present the fragmentation algorithm to solve the redundancy problem in multiple continuous queries that increases in the count or the amount of transmitting data in sensor networks. The fragmentation algorithm splits one query into more than two queries using the query index (QR-4ree) in order to reduce the redundant query region between a newly created query and the existing queries. The R*-tree should be reorganized to the QR-tree right to the structure suggested. In the result, we preserve 20 percentage of the total energy in the sensor networks.

• Korean Journal of Food Science and Technology
• /
• v.8 no.4
• /
• pp.236-241
• /
• 1976

A barley (20-30%)-wheat (80-70%) flour gave generally acceptable noodle-making characteristics, in which naked barley flour showed somewhat better results than covered barley flour. And also for over 40% barley flour in the barley-wheat flour, the addition of xanthan gum improved the noodle-making characteristics. A naked barley-defatted soy flour could make dried noodles with only high NSI (Nitrogen Solubility Index) defatted soy flour, however this mixture was not considered to be feasible for noodle-making. In cooking characteristics of barley-wheat flour noodles, naked barley flour was more acceptable than covered barley flour and xanthan gum influenced the water absorption and volume expansion of noodles during cooking. The firmness, cohesiveness and gumminess in cooked noodles made of the barley-wheat flour increased and then decreased as the amounts of barley increased. The mixing of defatted soy flour (high NSI) to naked barley flour increased the firmness and gumminess of cooked noodles.

• Korean Journal of Remote Sensing
• /
• v.23 no.6
• /
• pp.521-528
• /
• 2007

The overall objective of this research was to investigate various combination of segmentation parameters and to improve classification accuracy of object-oriented classification. This research presents a method for evaluation of segmentation parameters by calculating Moran's I and Intrasegment Variance. This research used Landsat-7/ETM image of $11{\times}14$ Km developed area in Ansung, Korea. Segmented images are generated by 75 combinations of parameter. Selecting 7 combinations of high, middle and low grade expected classification accuracy was based on calculated Moran's I and Intrasegment Variance. Selected segmentation images are classified 4 classes and analyzed classification accuracy according to method of objected-oriented classification. The research result proved that classification accuracy is related to segmentation parameters. The case of high grade of expected classification accuracy showed more than 85% overall accuracy. On the other hand, low ado showed around 50% overall accuracy.

• Journal of Korea Society of Industrial Information Systems
• /
• v.23 no.6
• /
• pp.29-38
• /
• 2018

In this paper, we developed a virtual cloth try-on (VTON) technology that segement the cloth image worn on the mannequin and applies it to the user 's photograph. The two-dimensional image-based virtual wear study which does not require three-dimensional information of cloth and model is of practical value, but the research result shows that there are limitations of of the current technology for the problem of occlusion or distortion. In this study, we proposed an algorithm to apply the results obtained from the DNN- based segmentation and posture estimation to the user 's photograph, assuming that the mannequin cloth reduces the difficulties in this part. In order to improve the performance compared with the existing one, we used the validity check of the pre-attitude information, the improvement of the deformation using the outline, and the improvement of the divided area. As a result, a significantly improved result image of more than 50% was obtained.