I am a microscopy image analysis noobie.It'd be great to have a discussion on this question.We have multi channel fluorescence images of dendrites and other brain structures obtained before and after an experiment on mouse.In the images obtained after the experiment,in some channels,fluorescence from 2 different viruses are present.Is there a way to separate the two fluorescence?
Hi Meen,
you can try to use multiple tracks for the laser set up instead of using serial track for the excitation and emission collection.
Hope this would help.
I'm assuming that the channels are separate images using different dyes but that the emission spectra are close enough that your filter sets can't completely block the undesired light. You can threshold out the lower intensities on each channel assuming that these are crosstalk from other dyes. You can then either use this as a mask to spatially filter each channel or as the new modified channel with reduced dynamic range. The effect of this is that you ignore all channel information below the chosen threshold. If you are using dyes that are too close in excitation and emission, then there is no way to separate the channels. It is important to choose dyes with fluorescence characteristics and appropriate filter sets that allow channel separation at the beginning.
In simple cases there is the possibility to separate the information channels by mathematical postprocessing. If for instance your measured intensity AM contains 100% information of true fluorescence distribution A and 10% of fl. dist. B, and measured channel BM contains 100% of fl. distr. B without crosstalk, then you can calculate A=AM-0.1*BM. In the general case it is a linear system of equations that can be solved.
Nonlinear effects (saturation..) and noise will limit the purity that can be achieved.
- Badges
- Science method