http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
오늘 본 자료
피부 질환의 확인: 전이학습을 이용한 백반증 심각도 측정
권영호 한국정보과학회 2023 정보과학회논문지 Vol.50 No.1
백반증은 표피에서 멜라닌 색소가 소실되어 발생하는 흔한 후천성 피부 질환이며 임상적으로 몸에 창백하거나 흰색 반점이 나타난다. 백반증은 조기 치료가 중요하지만 백반증은 통증이나 건강 문제를 일으키지 않는다. 따라서 백반증 환자는 피부 병변이 외부로 보일 때 치료한다. 백반증은 피부과 전문의의 주관적인 판단으로 치료되며, 의료 영상을 얻기 어렵기 때문에 영상을 통한 정량적, 객관적인 분석 방법이 제시되지 않았고, 소수의 의료 데이터 영상을 통해 많은 진단 방법이 개발되었다. 본 논문에서는 백반증 의료 데이터 수집의 한계를 극복하기 위해 전이 학습을 이용하여 영상 분할을 통해 백반증 부위를 진단하는 방법을 제안한다. 전이학습 모델은 U-net, FCN, Deeplab 등의 딥러닝 모델에 적용 가능성을 실험하여 선정하였다. 또한, 백반증 중증도를 측정하고자 RGB 피부 영상으로 변환하여 의료 현장에서 사용되는 VASI 점수를 이용하였다. 실험 결과, 불균형 백반증 이미지 데이터로 훈련했을 때 Deeplab에서의 f1-score와 IoU 점수가 기존 이미지 처리 방법 뿐만 아니라 U-net보다 우수했다. 또한, 이 논문에서 제안된 백반증 이미지에서 VASI 점수를 계산하는 방법은 백반증의 진단에 적용될 가능성을 보여 주었다.
Media를 충전한 간섭포기식 침지여상에 의한 흐름특성에 관한 연구
權泳浩,梁相鉉 全北大學校 1994 論文集 Vol.37 No.-
The flow pattern characteristics of submerged biofilters filled with three kinds of filter media were experimentally examined, and obtained results are as follows. 1. Flow patterns in submerged biofilters filled with three kinds of filter media was nearly complete mixing type when the parts of effluent are recycled, and nearly plug flow type without recycle. 2. Dispersion numbers of the flow recirculated was appeared 0.1352-0.2008 for A media 0.1156-0.1799 for B media and 0.1436-0.2035 for C media, respectively. Dispersion numbers of the flow not recirculated was appeared 0.0257-0.0614 for A media. 0.0122-0.0437 for B media and 0.0428-0.0762 for C media, respectively.
하악전돌증환자의 악교정수술후 하악각변화에 관한 임상적 분석
권영호,장현중,이상한 대한악안면성형재건외과학회 2000 Maxillofacial Plastic Reconstructive Surgery Vol.22 No.2
Predictional study for lateral change between pre- and post-orthognathic surgery has been emphasized mainly on anterior area of lateral profile ; upper lip, lower lip and chin et al. So interest for lateral profile change has been less in posterior area of lateral profile and literature analyzing gonial angle change is rare. The purpose of this study is to make prediction for gonial angle change possible and to offer somewhat treatment guidance for gonial angle to be improved by investigating overall gonial angle change between pre- and post-orthognathic surgery and inquiring into factors influencing on pattern of genial angle change. For this study 35 patients were selected retrospectively. Lateral cephalometric radiographs were taken in just pre-op time, pod 1 day, pod 1 year. They were analyzed and genial angles were measured. The results were as follows : 1.Gonial angle at pod 1 day was decreased about 9.3˚than pre-op and gonial angle at pod 1 year was increased about 4.0˚than pod 1 day. So genial angle at pod 1 year was decreased about 5.3˚than pre-op genial angle(p<0.01). 2.Mean pre-op gonial angle was 129.4˚, showing significantly high value than normal and mean gonial angle at pod 1 year was 124.1˚, showing value near to normal. 3.Mean gonial angle change between pre-op and pod 1 year was decreased about 5.4˚ in female and 5.3˚ in male. There was no statistically significant difference between male and female(p>0.05). 4.Principal factor influencing on decreased gonial angle in gonial angle change between pre-op and pod 1 year was amount of mandibular setback. 5.Principal factor influencing on increased gonial angle in gonial angle change between pod 1 day and pod 1 year was % horizontal relapse, and it was thought that resorption and bone remodelling on posterior area in mandibular distal segment also were related to increased gonial angle. 6.It is thought that sagittal split ramus osteotomy in mandibular prognathic patients with high value of gonial angle is effective to improvement of gonial angle, and In patients who have normal range of gonial angle and are required with excessive mandibular setback, short lingual cut method, additional resection of posterior margin of distal segment, Obwegeser II method will be considerd. 7.More prudent operation and careful post-op management will be responsible for maintenance of post-operactive stable gonial angle.