Digital Image Analysis: Cell & Nuclei Count


Counting cells




Image analysis yields reliable results when manual counting is too time-consuming in practice.

For example, a confluent monolayer (left image) was analysed by segmentation (main steps: Otsu-thresholding and watershed segmentation) and subsequent particle count (object on edges excluded). There were 653 cells counted within some seconds running an appropriate macro.monolayer_analyzed




The analysis result is shown in the second image as indicated by outlining the identified objects.

Obviously, some cells were not properly identified. (example indicated). The overall error is, however, below 3% as determined by counting the cells manually (yes, you can).