The Study Of The Tracking Algorithm Of The Neural Stem Cells

Stem cells are a group of unclassified cells and have two charactistics. One is that it can divide and multiplicate cells the same as its own. The other is that it can grow up some body cells which have some special functions. Neural adult stem cells still has these two charactistics. How to find out the law of the neural adult stem cells divide and multiplicate, and then culture some special ancestor cells to make them grow up some special neural organs. The final aim is to cure some neural disease by organ transplantation, which becomes difficult and popular more and more in the world.This dessetation is an attempt to solve this problem on the base of digital images analystics. The series images source is captured by digital microsopy imaging instrument II type. The cultured alive cells are derived from the neural stem cells of the brain of adult rats. The microscopy images are captured every 15 minutes. And one series is made up of 70 frames in average. The activaty status of the stem cells is shown on each frame. In order ro keep the cells alive, they are all cultured in unstained circumstances which have brought out relatively difficulties in image segmentation then also in tracking.The segmentation algorithm presented by the researchers of the Center of Image Analysisis of Uppsala University is introduced here. This algorithm is called region-based segmentation where seeds representing object and background pixels. The seeds are then used as a starting-point for watershed segmentation of the gradient magnitude of the original image. Over-segmented objects are thereafter merged based on the gradient magnitude, intensity, edge and shape between the adjacent objects. On the base of the result of the segmentation algorithm, an automatic tracking method is presented. In order to get complete and correct tracking results as possible the user can interactivelyverify and correct the crucial starting segmentation of the first frame and inspect the final result and correct errors if necessary. All cells are classified into active, inactive, divided and clustered cells. Different algorithms are used to deal with the different cell categories. A special backtracking step is used to automatically correct for some common errors that appear in the initial forward tracking process. The stemcells of two series images altogether more than 140 frames are tracked. The trace figure is used to show the tracking result. After evaluated by the human sights, the results are correct.