I have been using axiovision software (from Zeiss) for imaging my fluorescent strains. I wished to know as to what will give me an accurate quantification of pixel intensity? I have tried both auto live and fixed exposure time for imaging. To me, it appears like autolive gives image like how I see it with my eyes, but the exposure time with autolive option changes for every image. Fixing the exposure time, gives me images that may not be truly representative of how I see them through the eyepiece. Now, since I wish to perform quantitative analysis of these images, I wanted to know what is the best way to go about it. Does autolive or fixed exposure time work best or is there any other alternative? If anyone has experience doing similar stuff and could share it with me, it would be of great help.
This is actually a quite complex problem. Your eyes are lying to you, use quantitation with fixed exposure times. Standardizing your exposure times will allow you to compare fluorescent intensity between samples, genotypes, and conditions, something that is rendered useless by a variable exposure time. Use the range indicator software to adjust the black level and max levels to ensure you are in the linear range of the camera/CCD/PMT for wild type then apply those conditions methodically through your experiments. If your treatment or mutant conditions leads to brighter fluorescence, you should scale back the exposure time or light intensity to ensure you are accurately using the range of the detector. Quantitate at 12-bit (good) or 16 bit (preferred). There is no reason to use 8-bit greyscale with the current availability of cheap memory.
What light source are you using? Is it a Hg vapor arc lamp? If so, they can be quite variable in illumination brightness during the life of the bulb, and can even fluctuate during an experiment, so you should make sure you collect wild type or control data along side any treatment conditions. If you are using a laser-scanning confocal, or newer bulb sources, then light fluctuations are less of a problem and you can be reasonably sure the light levels do not drift too much on subsequent days. Still, always collect data for wt and treatment or whatever on the same days. You can buy fluorescent beads to image if you want to include an internal standard for fluorescent intensity.
Finally, are your transgenes integrated? If you’re using mosaic, extrachromosomal transgenes, your expression could be quite variable, further complicating the analysis.
Hi,
I agree that you need to take fixed exposure times, but one issue we have also found is that in some imaging software packages, the display image (and resulting tiff) are auto adjusted. So even if you took your images at the same exposure settings, the display (and tiff you would export) are set to crank up dimmer pixels. In that case, we manually set the lowest and brightest pixel points to be the same for each image. I think with Axiovision, if the display is set to linear, this shouldn’t be a problem, but its some thing to check.
Amy
I am also working with a zeiss microscope. I had the same question at the begining even if I am not doing quantification with my pictures. What works best for me is to save the pictures in .zvi and export them in ImageJ to analyse, merge or do anything with it. Like this you should not lose any information for your quantification. I now use fixed times that work for almost all images.
I hope it helps !
Thanks for all your responses. It helps me greatly. AKW, could you tell me specifically tell me how you set brightest and lowest pixel points in axiovision(if thatis the software you are using)?