The x and y coordinates of the cell

We used the Cellprofile to analysis the image data that included three channel Nucleus, Cytoskeleton and Protein Expression. The cell profiler output data included quantitative parameters for each cell represents the x and y coordinates of the cell, the area, the class which the cell belongs to etc. When we manually check the x and y coordinates of the cell in each matched image, we found it did not exactly reflect the location. I am wondering if there is any suggestion?




May I ask how you manually checked the x,y coordinate ? By which tool/software and in what unit was the manual check?

The x,y coordinate reported by CP is the center of an object, in pixel unit. Is it possible that you check the edge or a different center? or in different unit?

An experiment to test: make a crop of the image that contain a single object, where object’s XY is also the center of the cropped image. Then re-test your method vs. CP report.


Depending on what measurement you’re checking, also, the center reported can change- Location_Center_X and Location_Center_Y are the “original” calculated center, but others reported by MeasureObjectSizeShape, MeasureObjectIntensity, or other modules actually are the center of mass or center of intensity of the object, so be careful which measurement you are checking. You can check the help for each module for a further discussion of exactly what each reported measurement is.

(You don’t mention a Z plane, but there also was a bug in 3.0.0 on 3D images that switched recording the X and Y coordinates; that got fixed last week and you can download the fix in our next release).

Thank you so much for the quick reply during the weekend. We will check based on your suggestion and will update once we find the solution.

Best wishes,