Dear Gilles ,
I am having a Min error not converging problem . It appears to occure when the number of single cells ( that do not connect with other cells to form structures) is relatively large. It does change the results since the algorithm calculates many “false positive” networks (as can be seen in the sample I have attched) .Since I am ommiting the results of these images from the expriment, I am concerend that experimrntal conditions with relatively low tube-formation are “underpresented” and due to that the overall results are false. Should I reduce the size of the image in order to correct this situation (as you suggested) ? If so , how?
Dear Gilles ,
You already posted the very same thing on a different thread:
While it’s no problem to start a new discussion, you should link to previous related discussions so that others will understand the context and will be able to help you.
If you want to address specific persons on the forum, it helps to mention them using
@ followed by their user name: @carpentier. The forum software will then send them a notification.
Looking at your image, I’d suggest that you do some shading correction before starting the analysis. This will help to get more reliable results and might also help to avoid the mentioned MinError issue.
Thanks. Shading corrections did’nt help . Maybe I was doing it wrong , I have used the tools in adobe photoshop as well as the tools in imageJ .