I have some SEM images of nanofiber meshes with different densities and want to compare them. Therefor I would like to analyze the images to evaluate data statistically.
Dear Clemens,
First of all, what you mean by "density"?
I understood, that you would like to measure number of fibers per area unit. In my opinion, it is impossible to evaluate this kind of data from SEM images. However, you can use porosity instead of density of fibers. Some researchers evaluate porosity based on the free space between the fibers in the picture. Also, you can use some formulas to calculate porosity of your material by using density of polymer and mass of your sample:
http://doi.wiley.com/10.1002/mame.200900052
I hope I understand your question well, but if you will precise it, I might be more helpful.
Best regards,
MW
Thank you Michal.
Well let me explain: We get meshes send by another scientist and these meshes are each time a little different. I noticed that cells perform different on some of our meshes.
I noticed in SEM images that even so the fiber diameter is the same in all meshes, pores seem to be smaller and the amount of fibers per SEM image seems to be elevated. I need to statistically analyze this but therefor I need a way to measure pores vs fibers of all of these images.
Best regards
Dear Clemens,
So, as I mentioned earlier, you can use the image and calculate the area of fibers and area of pores in the first "layer" of fibers. It is quite easy, you can use ImageJ to do that. Then you just compare area of pores and area of whole picture. I heard about this kind of approach, but I never did it myself. I will try to get in touch with my colleague who used this method and maybe I will be able to send you some papers on that.
Best regards,
MW
Dear Clemens, the question you raise may be covered in our recent biomaterials paper, as well as the reference we've included. in any event, like Michal said…ImageJ is the key...
ich hoffe das hilft mal weiter…
Boris
Dear Clemens
Be aware that electrospun meshes can have 2 different sides. Quite often the side to the target is more dense then the side exposed to the air. Please don't understand this as critics, I realized already sometimes that researches are not aware of this and seed cells random on the different sides of the scaffold, hence increasing their variations.
And yes,it is difficult to avoid variation in the spinning result without climate control of the lab or the equipment, sometimes even with.
Regards
Marc
Hiya,
I agree with Michal Wojasinski. ImageJ would be a useful tool to use. If you have images in similar magnifications and all, you could try segmentation thresholding. Try on a several small regions, and then compare that to the full image. And repeat for whatever different kinds of meshes you have.
If you need porosity data, then you might be better off doing some kind of porosimetry analysis. But if you're limited to SEM images, then you can easily compare the area occupied by the fibre mesh per unit area in an image! However, I'm not so sure if you should evaluate "density" from 2D data.
I could have a look at one of your images for you =)
Best wishes.
Hi guys please take a look to this for fiber intersection density http://www.sciencedirect.com/science/article/pii/S0142961210004230 and figure 4 of http://www.sciencedirect.com/science/article/pii/S1742706113004595 I wrote the software for this analysis. Please let me know if you find this useful. Thanks, Antonio
I agree with those who assert that a 2D method like SEM is not a very accurate way to measure bulk density/porosity. Simple measurement of dimensions and weighing will give a fair answer (using material density), and using a densitometric kit on your 4 or 5 decimal balance, where you weight in air and when submerged in a liquid will give a more accurate one. We use one similar to the one shown in the clip: http://www.youtube.com/watch?v=ou_HJB6H6P0
this video might be useful for you.
https://www.youtube.com/watch?v=WxTu2fn7QxY
- Badges
- Science topic
- Similar topics
- Biological Science
- Molecular Cell Biology
- Culture Cells
- Cell Line