Gray values after color separation
Unlike immunofluorescence, the colors of markers in immunohistochemistry images are contained in one image – in order to ease analysis HistoQuest performs a color separation.
Color shades are picked manually and optimized by manipulation of specific values. Optimization results are displayed in near real time.
The automatic color separation produces a gray value channel image for each marker, allowing treatment with the application of TissueGnostics algorithms for single cell identification.
In HistoQuest, reference shades are selected semi-automatically. RGB color values are manually chosen with a color selection tool. An automatic pre-segmentation simplifying color picking is available.
The color shades acquired in this way can be adjusted and optimized for automatic nuclear segmentation by manipulating the gain, the color space and blur. The values in color adjustment are calculated very rapidly with interim results.
Nuclear segmentation
Automatic nuclear segmentation is achieved via TissueGnostics patented set of algorithms.
Once rapid test calculations done on representative images have established a good segmentation, nuclear segmentation as well as the measurement masks can be automatically computed for virtually any number of corresponding images.
Nuclear segmentation in HistoQuest is completely automatic after the input of a few starting values:
- Average nuclear size
- Discrimination by area (Exclusion of smaller nuclear sections)
- Discrimination by gray value (Exclusion of weakly stained nuclei)
- Background threshold (Default setting is automatic)
