Development of quantitative lmage analysis library for biomedical imaging applications생체 영상 어플리케이션을 위한 정량적 영상분석 라이브러리 개발

Cited 0 time in webofscience Cited 0 time in scopus
  • Hit : 298
  • Download : 0
As imaging technologies are advanced, a lot of biomedical images were generated from imaging devices such as microscope, CT, MRI, PET, and SPECT. Quantitative analysis can analyze massive quantities of biomedical images and also duplicate the previous experimental results. Most of the commercial software products provide only a few functions to satisfy their specific objectives. Thus, we sought to develop a quantitative image analysis library to fulfill all the requirements of imaging modalities. Developing the library needs to have various functions for object segmentation, feature extraction, and classification to fulfill all the requirements of imaging modalities. The developed library contained multi-level adaptive thresholding and watershed transform functions as segmentation functions; 62 feature extraction functions from size, shape, intensity, skeleton, and convex-hull operation; erosion, alternating sequential filtering (ASF) as noise removal functions; conditional dilation, gray-scale reconstruction as reconstruction functions; image normalization and cubic surface fitting algorithm as image correction functions. Using this library, we developed three different types of applications for quantitative image analysis. We implemented functions to find apoptotic candidate genes and cell cycle related genes in 2D microscopy imaging. Seven apoptotic candidate genes were found that were already validated by western blotting and ELISA. Nine candidate genes related to cell cycle showed more than twice expression levels in various tumor tissues compared to normal tissues and need to be validated by biochemical experiments. The intensity and volume of brain tumor from 3D SPECT images were measured by using 3D automatic marker-driven watershed transform. The volume of mouse abdominal fat was measured by three level adaptive thresholding based on the intensity histogram in 3D MR images. Through the implementation of these applications, AdoIC has been further de...
Kim, De-Sokresearcher김대석researcher
한국정보통신대학교 : 공학부,
Issue Date
392953/225023 / 020064609

학위논문(석사) - 한국정보통신대학교 : 공학부, 2008.2, [ vii, 93 p. ]


Image Segmentation; High Content Screening; Quantitative Image Analysis; Marker-driven Watershed Segmentation; 마커기반 워터쉐이드 기법; 영상 분할; 하이 컨텐츠 스크리닝; 정량적 영상 분석

Appears in Collection
School of Engineering-Theses_Master(공학부 석사논문)
Files in This Item
There are no files associated with this item.


  • mendeley


rss_1.0 rss_2.0 atom_1.0