Safari Books Online is a digital library providing on-demand subscription access to thousands of learning resources.
- Create BookmarkCreate Bookmark
- Create Note or TagCreate Note or Tag
- DownloadDownload
- PrintPrint
Chapter XIX. Image Processing and Machin... > MICROARRAY SEGMENTATION METHODS
MICROARRAY SEGMENTATION METHODS
Image Processing Techniques
Fixed or Adaptive Circle Segmentation
Fixed circle segmentation is the earliest method developed for microarray image analysis. This algorithm is implemented by Eisen et al, (Eisen, 1999) and it is included in the ScanAnalyze software tool. The method assumes that all the spots are circular with a constant radius. A circular mask of a fixed radius, called target mask, is placed on each spot location, considering all the pixels inside the mask as foreground pixels. On the other hand, background contains any external pixel which is close to the corresponding spot. Fig. 3 shows the way that masks are placed on each spot using the ScanAlyze software.
The elimination of the constant radius assumption was the next step in microarray image analysis studies, generating the adaptive circle algorithm. Assuming that the shape of all spots is circular, the radius for each spot is automatically estimated or manually adjusted by the user, for each spot. For instance, Dapple (Buhler et al., 2000) estimates the radius of the spot using the Laplacian-based edge detection. The manual approaches are extremely difficult and time consuming due to the large amount of microarray spots contained in a single image.
