返回Google圖書搜尋

Image analysis system for rapid data processing in Tradescantia-micronucleus bioassay

Xudong JongGuangheng LinJianhua XuWenjie XiaWeichi SunTe-Hsiu Ma

出版	American Society for Testing and Materials, 1990
URL	http://books.google.com.hk/books?id=NO94zgEACAAJ&hl=&source=gbs_api

註釋Tradescantia-micronucleus (Trad-MCN) bioassay is a well established short-term test for environmental mutagens in liquid or gaseous forms. For large-scale experiments that involve hundreds of slides, the micronuclei scoring process, data collection, and analysis are labor intensive and time consuming. Using computer image analysis process, a rapid data processing system was developed. This new facility is called the Trad-MCN Image Analysis System (Trad-MCN-IAS). The hardware of this system is composed of a photomicroscope attached to a video camera and a US-386 computer. The software for this system was written in FORTRAN language and run under an FG-105 real-time image processor. The operation begins with multi-threshold image segmentation, followed by mathematical morphology-based micronucleus extraction, and judging of normal tetrads with pseudo-color display. Using this system, a large number of tetrads (the four-cell stage of the meiotic pollen mother cells) image on the screen can be scored simultaneously for normal and micronucleus-bearing tetrads. The number of micronuclei in a large population of tetrads can be recorded and put in memory for data analysis. For high efficiency, four microscopes may be installed on a single Trad-MCN-IAS to gather tetrad images and count micronuclei; therefore a large number of samples can be scored and the experimental data processed in a short time. This system has several advantages: non-bias in the scoring process, automation, and minimizing the variance of scores introduced by different human observers.