Track Objects: merge identified objects

Dear all,

first of all thanks to the developers for the great job on CellProfiler. This software is awsome!
I come from Germany, therefore I apologize for any bad english.

I am using a CellProfiler pipeline to identify cells in white light images. The cells have been imobilized and are tracked over 7 days. During this period of time, cells either
I) divide while the “area” they occupy increases, or
II) remain in same conditions with same sizes, or
III) increasing their volume, also increasing the area they occupy.

Since the cells are not in a real monolayer and due to general issues regarding white light images, the divided cells cannot be properly tracked by CellProfiler. However, I am only interested in the object area they occupy. I think, there is a significant difference between those 3 possibilities I discribed above.

I am running the following pipeline using a movie acquiered from images of 6 days as the source:

LoadImages
ColorToGray
FindEdges
IdentifyPrimAutomatic
MeasureObjectAreaShape
TrackObjects
OverlayOutlines
PlaceAdjacent
SaveImages
ExportToExcel

Now it comes to the problem. When I run this pipeline, the TrackObjects module usually divides the track objects when they overcome a distinct size (fig1)

http://img168.imageshack.us/img168/8539/tracked6d.jpg
Fig1: Sorry for the poor quality. Divided object can be seen in the 6th image in the upper right part for example.

Is there any possibility to merge those objects so that I can simply use the output data of any objects to compare the data?

Thanks in advance for any answers,

Karsten

Hi Karsten,

The performance of the TrackObjects module is dependent on the objects given as input. TrackObjects does not actually divide the objects; any segmentation problems with objects originates with the identification modules. So if your objects are being erroneously divided, I would suggest looking to the IdentifyPrimAutomatic module in your pipeline and adjust the settings in order to get the cell segmentation that you want. From the looks your images, you may want to start with increasing the smoothing filter size and/or the maxima separation distance.

Hope this helps!
-Mark