Aligning grids derived from fluorescent images of an array of signal sources

Abstract

Disclosed are methods of processing images, and in particular aligning and orienting grids from a fluorescent array of signal sources, by linking nearest neighbor sources to form a hexagon, and then using the hexagon lines for alignment and orientation of the grid. Also disclosed are methods and algorithms for aligning image with grid and correcting for signal beads which are smaller than a grid field and the shift in a grid field. These methods can be used where the signal sources are fluorescent images from a microarray. Also disclosed are methods of automated watershed clustering following transformation.

1 . A method of aligning grids derived from fluorescent images of an array of signal sources in different color channels, where the grid lines separate the signal sources, with a grid derived from a fluorescent image of the array of signal sources in a first color channel, where each signal source is composed of several pixels, comprising:
 examining the difference between pixels which are closest to the edges of the array in the first color channel to find a series of peak intensities representing the respective edges of the array;   determining the respective edges of fluorescent images from different color channels, by examining, for said different color channels, the difference between pixels which are closest to the edges of the array to find a series of peak intensities representing the respective edges of the array; and   aligning the respective edges of the array from said different color channels with the respective edges of the array in the first color channel.   
   
   
       2 . The method of  claim 1  further including the steps of finding the rotation angle of the array of signal sources in the first color channel following finding the respective edges of the array; and
 rotating the arrays from said different color channels by aligning their respective edges with the edges of the array in the first color channel.   
   
   
       3 . The method of  claim 1  wherein the signal sources are microbeads. 
   
   
       4 . The method of  claim 3  wherein the microbeads are labeled with a fluorescent dye. 
   
   
       5 . A method of analyzing a fluorescent image of an array of signal sources using a grid filed which divides said signal sources, wherein said signal sources are smaller than a respective surrounding grid field and can be expected to shift position within said surrounding grid field and thereby prevent accurate reading of the signal sources, comprising:
 finding the position of individual signal sources in individual fields in the grid by establishing a window in respective said individual fields, where said windows have a smaller area than the field they are respectively established in;   establishing pixels for the image where the pixels are sized such that there are several pixels in each window;   moving the windows to several different positions in their respective fields and determining the mean intensity of the pixels within the windows; and   determining that the window is directly over the actual bead position when it is at the location of maximum mean pixel intensity.   
   
   
       6 . The method of  claim 5  wherein the respective windows are not placed over the extreme edges of the respective grid fields, in order to avoid regions of excess noise. 
   
   
       7 . The method of  claim 5  wherein if the bead intensity in a grid field is weak, the mean intensity at the center of the grid field is used as the actual intensity. 
   
   
       8 . The method of  claim 7  wherein the bead intensity in a grid field is deemed to be weak if it is below a threshold for signal source intensity.