496 diame mounted with ProLong Gold reagent with 49,6 diamidino 2 phenylindole . Automated Estrogen Receptor Pathway Image Acquisition and Analysis Images were analyzed using algorithms that have been described. Tumor was distinguished from stromal elements by cytokeratin signal. Coalescence of cytokeratin at the cell surface was used to localize cell membrane cytoplasmic compartment within the tumor mask, and DAPI was used to identify the nuclear compartment within the tumor mask. Targets were visualized with Cy5, this wavelength is used for target labeling because it is outside the range of tissue autofluorescence. Multiple monochromatic, high resolution grayscale images were obtained for each histospot using the 106objective of an Olympus AX 51 epifluorescence microscope with automated microscope stage and digital image acquisition driven by a custom program and macrobased interfaces with IPLabs software.
Images for each histospot were individually reviewed. Two images were captured for each histospot and for each fluorescent channel, DAPI, Alexa 546, and Cy5, one image in the plane of focus and one 8 ?`m below it. The compartmentalization and quantification of the target protein signal within each pre defined compartment for each histospot was performed as follows. GSK461364 First, the Alexa 546 signal representing cytokeratin staining was utilized to generate an epithelial cell mask that excludes all other stromal elements. This signal is binary gated in order to identify whether a pixel is within the tumor mask or not, all white pixels are part of that mask and all black pixels are not part of this compartment.
Similarly, the nuclear compartment is defined as pixels that demonstrate DAPI staining within the plane of focus and within the region defined by the tumor mask. The DAPI image is also binarized to generate a mask of all nuclei within the sample by subtracting out overlapping pixels with the cytoplasmic mask, all white pixels are part of this mask while all black pixels are not. To ensure that only the target signal from the tumor and not the surrounding elements is analyzed, the RESA PLACE algorithms were utilized. The RESA algorithm provides an adaptive thresholding system. In general, formalin fixed tissues can exhibit autofluorescence and sometimes analysis can give multiple background peaks. The RESA algorithm establishes the predominant peak and then sets a binary mask threshold at a slightly higher intensity level.
RESA eliminates all out of focus information by subtracting a percentage of the out of focus image from the in focus image, based on a pixel by pixel analysis of the two images. This eventually allows more accurate assignment of pixels of adjacent compartments. Finally, we utilize the PLACE algorithm to assign each pixel of each image to a specific subcellular compartment. All pixels that cannot be accurately assigned to a compartment with a degree of confidence of 95 are ultimately excluded. Additionally, all pixels for which intensities are too similar in t