http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
골파괴를 동반한 부비동 점액낭종의 근치수술후 동시재건술 치험례
임 풍,이종원,조성필,강윤섭 大韓成形外科學會 1991 Archives of Plastic Surgery Vol.18 No.5
Mucocele is a chronic, expansile & cystic lesion of the paranasal sinuses, which contains sterile, mucoid secretions. Although the etiology and pathogenesis of a sinus mucocele are uncertain, the accepted theory points to an obstruction near or at a sinus outlet that causes retention of mucous secretion within the sinus cavity. Osseous trauma, tumors or a mucosal thickening (chronic inflammatory disease) in the areas of outlet are the major associated factors. The mucoceles are most frequently found in the frontal sinus. Progressive enlargement of mucocele may cause local bone destruction, variable functional impairments, and facial deformities. In this case, simple surgical removal of mucocele may produce severe facial deformyty and functional impairment postoperatively. To minimize postoperative facial deformities, we have applied primary reconstuctive procedures, utilizing cranial bone grafting for postoperative defect in 2 patients. Judging from postoperative cosmetic and functional viewpoints, the results were fairly satisfactory.
강윤섭(Yoonseop Kang),최승진(Seungjin Choi) 한국정보과학회 2010 정보과학회 컴퓨팅의 실제 논문지 Vol.16 No.4
소셜 네트워크를 비롯한 네트워크로부터 커뮤니티를 발견하려면 네트워크의 노드를 그룹 내에서는 서로 조밀하게 연결되고 그룹 간에는 연결의 밀도가 낮은 그룹들로 군집화하는 과정이 꼭 필요하다. 군집화 알고리즘의 성능을 위해서는 군집화의 기준이 되는 유사도 기준이 잘 정의되어야 한다. 이 논문에서는 네트워크 내의 커뮤니티 발견을 위해 유사도 기준을 정의하고, 정의한 유사도를 유사도 전파(affinity propagation) 알고리즘과 결합하여 만든 방법을 기존의 방법들과 비교한다. Finding communities from network data including social networks can be done by clustering the nodes of the network as densely interconnected groups, where keeping interconnection between groups sparse. To exploit a clustering algorithm for community detection task, we need a well-defined similarity measure between network nodes. In this paper, we propose a new similarity measure named "Common Neighborhood Sub-graph density" and combine the similarity with affinity propagation, which is a recently devised clustering algorithm.
강윤섭(Yoonseop Kang),최승진(Seungjin Choi) 한국정보과학회 2010 한국정보과학회 학술발표논문집 Vol.37 No.2C
최신의 비지도 학습 신경망 모델인 볼츠만 기계는 모델의 유연성과 실용적인 훈련 시간 등의 장점으로 패턴 인식 및 인공지능의 많은 분야에 응용되어 사용되고 있다. 그러나 볼츠만 기계의 시간복잡도는 데이터와 모델 크기에 정비례하여 이를 감소시키기가 어렵다. 이 논문에서는 볼츠만 기계의 시간복잡도를 줄이기 위하여 GPGPU 기술을 활용한 알고리즘 병렬화 기법을 제안하고, 알고리즘을 실제 데이터에 적용하여 알고리즘 설정에 따른 학습에 들어가는 시간을 측정하여 CPU를 사용하는 알고리즘과 비교하였다. 실험 결과, GPU를 사용하였을 때 CPU를 사용했을 때보다 실행 시간이 크게 단축되었다.
화상을 입힌 생쥐의 흉선과 비장 임파구아형과 조직학적 소견의 변화
한기택,안상태,강윤섭,임 풍 大韓成形外科學會 1991 Archives of Plastic Surgery Vol.18 No.4
Due to various documented abnormalities of cell mediated immunity, sepsis still remains the major cause of morbidity and mortality after initial survival from extensive thermal burn. However, the exact mechanisms of immunosuppression after extensive thermal burn remains uncertain. For th observation of immune state, changes of lymphocyte subpopulation have been usually measured in lymphocyte samples from peripheral blood by many reports, but it is only a small fraction of the total lymphocyte pool in the body, and may not present valid picture of the immunological state of the individual. Therefore, analysis of lymphocyte phenotype in thymus and spleen is thought to be necessary in understanding more exact mechanism of immunological changes in extensive thermal burn. The aim of this study is to evaluate the changes of immunocompetency by observing the changes of subpopulation of spleen and thymus lymphocyte using monoclolnal antibodies and changes of tissue with light microscope in burned mice. Cell populations were isolated from spleen and thymus of control and burned mice on the 1st, 3rd, 5th, 7th, 14th and 21st postburn days(PBD). The cell populations were measured with monoclonal antibody by flow cytometry. The results were as follows: 1.In thymus, Thy 1.2+, L3T4+ and LYt2+ cells decreased markedly on the 3rd PBD. Thy 1.2+ and L3T4+ cells decreased also on 5th, 7th, 14th PBD and returned to control level on 21st PBD, but Lyt2+ cells increased more significantly than control after 3rd PBD and returned to normal on 21st PBD. Qa2+ cells revealed most significant reduction on 7th PBD. 2.In spleen, Thy 1.2+, L3T4+ and Ia+ cells increased on 3rd PBD and then decreased on 7th and 14th PBD. Lyt2+ cells increased in all the burned groups. Qa2+ cells increased on 3rd, 5th PBD and returned to normal level. L3T4+ cell/Lyt2+ cell ratio revealed signicant reduction, but they recovered by the 21st PBD. 3.In histopathologic findings of the spleen & thymus, the all cell subpoulations decreased at 1st PBD, but more increased than that of group of 3rd PBD. And then they revealed normal architecture after 5th PBD.