Hi, I'm interested in using imageHTS and associated libraries for some screening data. My only hesitation at the moment is the amount of work / specialist image processing knowledge (of which I have ~0) involved in writing cell segmentation and feature extraction routines. The only example I can find is the built in segmentATH which the ImageHTS vignette describes as 'specifically designed to segment cells stained for DNA and cytoskeletal proteins' (not appropriate for my data) and the sections at the end of the EBImage vignette - which look suspiciously simple compared to segmentATH(). Before I dive into writing my own segmentation routine I just wanted to check that I'm right in thinking there isn't another library of pre- made routines or any other good resources for doing this in R/Bioconductor? -- Alex Gutteridge

Dear Alex I am not aware of any shortcuts. Note that the segmentation need not necessarily be 'true' (compared to the real cell or nucleus boundaries) as long as the subsequent feature extraction leads to useful (reproducible and discriminating) quantitative features. Best wishes Wolfgang Il Dec/6/10 4:00 PM, Alex Gutteridge ha scritto: > Hi, > > I'm interested in using imageHTS and associated libraries for some > screening data. My only hesitation at the moment is the amount of work / > specialist image processing knowledge (of which I have ~0) involved in > writing cell segmentation and feature extraction routines. The only example > I can find is the built in segmentATH which the ImageHTS vignette describes > as 'specifically designed to segment cells stained for DNA and cytoskeletal > proteins' (not appropriate for my data) and the sections at the end of the > EBImage vignette - which look suspiciously simple compared to segmentATH(). > Before I dive into writing my own segmentation routine I just wanted to > check that I'm right in thinking there isn't another library of pre- made > routines or any other good resources for doing this in R/Bioconductor? >