Image analysis with Immuno Membrane

The purpose of this program is to measure the number of stained membranes of the cells identified within an image. The typical application for this algorithm is the quantification of HER2 staining.

The protocol :

The different stages of image analysis using the Immuno Membrane algorithm are as follows :

  • Identification of the color plane
    The algorithm carries out automatic color deconvolution (DAB/H or AEC/H). This deconvolution can subsequently be adjusted by the user.
  • Image segmentation
    Image segmentation enables identification of the set of nuclei that is to be analyzed then identification of the membrane pathway. This process is carried out automatically and can subsequently be refined using a cursor.
  • Definition of staining positivity thresholds
    The segmentation of objects can be refined via morphomathematical processing. The thresholds of the four categories of staining intensity are defined.

Results are expressed in the number and percentage of objects for each of the four categories then compared with the total number of objects present in the region of analysis. Analyzed objects can also be individualized and described on a one-by-one basis.

Nuclei are categorized according to their staining intensity and the completeness of the membranes surrounding them. They are distributed into four categories, from negative to intensity 1+, 2+ or 3+. The proportion of nuclei surrounded by a positive and complete membrane, out of all the detected nuclei, which is the ratio of the number of stained nuclei against the total number of nuclei, then determines the staining index.