Digital Image Analysis: Cell & Nuclei Count


Counting nuclei



For counting nuclei, image analysis followed the same principles as already described. Background correction (if necessary), segmentation and particle analysis.

For optimizing the results, some additional steps (e.g. filtering in the spatial or frequency domain) may be included.

The process runs automatically when implemented in a marco.








The analysis results are presented as outlines generated from an image mask. For the example above, 710 nuclei were automatically detected. When nuclei were counted manually, 722 objects were identified Therefore the overall error was less than 2%. Three weakly stained nuclei were not detected by the programm (false negative), one count was artificial ( false positive) and some overlapping nuclei were not correctly separated (encircled in red, for example).