DC Field | Value | Language |
---|---|---|
dc.contributor.advisor | Ye, Jongchul | - |
dc.contributor.advisor | 예종철 | - |
dc.contributor.author | Huh, Jaeyoung | - |
dc.date.accessioned | 2021-05-12T19:33:46Z | - |
dc.date.available | 2021-05-12T19:33:46Z | - |
dc.date.issued | 2020 | - |
dc.identifier.uri | http://library.kaist.ac.kr/search/detail/view.do?bibCtrlNo=909925&flag=dissertation | en_US |
dc.identifier.uri | http://hdl.handle.net/10203/283844 | - |
dc.description | 학위논문(석사) - 한국과학기술원 : 바이오및뇌공학과, 2020.2,[v, 37 p. :] | - |
dc.description.abstract | In this study, we propose the method for image reconstruction in optical diffraction tomography (ODT). Because of the physics of ODT, there is a missing cone part in the reconstructed images using the conventional method. It generates many problems. The first is a lower refractive index (RI) problem. It means that the cell intensity cannot reflect the original cell intensity. The second is the cell elongation problem. If we see the plane of the image, two-direction plane images have elongated cell image along the missing part direction. So, it cannot reflect the original cell image. There are some algorithms to address this problem. However, it takes a very long time to generate an image or it makes artifacts. Therefore, in this study, we apply the deep learning method to solve the missing cone problem with avoiding those disadvantages. | - |
dc.language | eng | - |
dc.publisher | 한국과학기술원 | - |
dc.subject | Deep learning▼aSupervised learning▼aOptical Diffraction tomography▼aMissing cone▼aMicroscopy | - |
dc.subject | 심층 학습▼a지도 학습▼a광학 회절 현미경▼a콘 모양 손실▼a현미경 | - |
dc.title | Deep learning-based optical diffraction tomography reconstruction | - |
dc.title.alternative | 심층 학습 기법을 이용한 광학 회절 현미경 이미지 복원에 관한 연구 | - |
dc.type | Thesis(Master) | - |
dc.identifier.CNRN | 325007 | - |
dc.description.department | 한국과학기술원 :바이오및뇌공학과, | - |
dc.contributor.alternativeauthor | 허재영 | - |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.