Posted by
Guenter Giese on
Nov 15, 2005; 10:24am
URL: http://imagej.273.s1.nabble.com/Colocalisation-analysis-tp3704464p3704466.html
Hi Kong,
I think this is caused by the type of conversion you use.
First split the 24bit RGB images into their channels and then use the
respective green or red channel for your analysis. Otherwise you reduce your
brightness (intensity) values due to the conversion process.
Hope this helps
Guenter
------------------------------------------
Dr. Guenter Giese
Light Microscopy Facility Manager
Dept. of Biomedical Optics, MPI fuer Medizinische Forschung
Jahnstr. 29, D-69120 Heidelberg, Germany
Phone (+49) 6221-486-360 (Fax: -325)
e-mail:
[hidden email]
http://lightmicro.mpimf-heidelberg.mpg.de-----Ursprüngliche Nachricht-----
Von: ImageJ Interest Group [mailto:
[hidden email]] Im Auftrag von Lee
Kong Heng
Gesendet: Tuesday, 15 November, 2005 10:40
An:
[hidden email]
Betreff: Colocalisation analysis
Hi Folks,
I think I'm doing something wrong as the scatter plot should filled
the whole graph instead just part of it:
http://img383.imageshack.us/my.php?image=funnyscatterplot2lm.jpgHere's what I did:
1. Open all files 24 bit TIFF (Green only and Red only fluorescence
images).
2. Convert to 8 bit for each image: Image>Type>8 bit
3. Analysis: Plugins>Colocalisation analysis>Manders coefficients
4. Manders Calculator appears:
Set green image to green channel, red to red channel
None for use ROI
Selected Display Colour Scatterplot
Selected Exclude zero-zero pixels
5. Press ok. Statistics seems fine but scatterplot is funny.
I'll be grateful for any advice that you may share and thank you.
Sincerely,
Kong Heng