Andor Microscopy School - Principles in Confocal Microscopy

I'm Claudia Florindo, and today I will give you an introductory talk about the principles in confocal microscopy. For the summary of today's talk, I will quickly review the fluorescence microscopy microscope, and I then will introduce what is a confocal microscope. Following, I will explain point scanner microscopes, the pinhole, what is the pinhole, the spinning disk confocal microscopes, and which microscope is the one to choose depending on the application you are going to use.

So as a quick review of the epifluorescence microscope. As I mentioned on the previous talk, we have the light source and the filter cube with the respective excitation, dichroic, and emission filter. The light comes through the light source, is reflected with the dichroic mirror, goes to the sample, and the sample reflects back the light passing through the dichroic now, the second barrier emission filter and being captured by the detector.

This would mean, and in this design, the detector is at the side and the light source is at the top. This would mean that in the sample, all the light that is in the sample from all the planes, the plane in focus and other planes that are not in focus, will be captured by the detector.

This schematic drawing shows what is captured in an epifluorescence microscope. We have, for example, the light coming from the focal plane in this Z section here in the middle of the cell, but the detector is also able to capture light and will capture light that comes from planes above and below the focal plane. What would actually want to become the image more sharp and then increase the signal-to-noise and the resolution of the image is to avoid the out-of-focus light and just capture the light that is in focus on the specific focal plane. And then to capture all the cells, we would need to find a way that the focal plane will go through all the Z section of the cell.

This is an example of a kidney tissues section in the widefield microscope, and this is what we would like to have avoiding this out-of-focus light, this haze that will be more significant as samples get thicker. This is a widefield image, and this is a confocal image.

But another problem is posed is that how can we acquire a 3D image in a 2D imaging system. And as you can see here, we have a 3D image of a Drosophila egg chamber. And, in fact, to do this 3D image, what we have done is that we have acquired 424 images. So we acquired in a plane, moved the stage a bit higher, acquired another plane and another and another and another, 424 in green and 424 in red.

If I organize all my 424 images side by side, you can see all the images that I have acquired to do this 3D image of the Drosophila egg chamber. I have also organized these 424 images in steps, 10 by 10. This is the 1st, this is the 10th, this is the 20th. You know why? That you can clearly see the Z step going through all the images that we have acquired. We can then visualize this 3D stack going through Z.

So, now, we are going through the Z of the 3D stacks of the Drosophila egg chamber and analyzing all the sections that I have acquired and exposed on the previous gallery.

And, now, I can use computer software to do the 3D reconstruction of the sample and analyze the sample in any direction, on the Z, on the Y, and visualize it through the inside. To do this, it's ideal, especially on a thick sample, to acquire the image with a confocal microscope, and to acquire with proper Z sampling in a way that when you go from one Z slice to the next Z slice, you have all the information you need to, later on, be able to reconstruct your 3D image.

So, in a way, a confocal image is like a book. So we have all the many pages of the book and each page shows a bit of information that you have available to read as I showed you in the previous images. Reading a page in a book gives you an idea of the information on that page but doesn't give you the full view of the book. So to see the whole book, to see the whole sample, to see the whole cell, you need to acquire all the slices, you need to read all the pages of the book, and then do your 3D reconstruction.

And the question now is how is a confocal image formed. So, like, how do we remove the out-of-focus light in a way that then you can do proper 3D reconstructions and visualize quite well inside your sample?

For acquiring a confocal image, we have this Drosophila embryo, the specimen. It's actually too thick for conventional fluorescence microscopy. This is actually the region that I want to observe, a region inside the embryo.

If I have the light here and I focus the light into my region of interest, goes into the objective, the confocal has a pinhole that will allow the in-focus light to be captured through the detector. But what happens if... So this is the in-focus light. What happens to the rays of light that go here or here? Will they also be captured by the detector?

This was the region of interest as I showed you before, and it passed through the pinhole, and it was in the detector, but if there is fluorescence coming from another region that it's not in focus when passing through the objective, it will not be focused then to be able to reach the pinhole and to pass through the detector, so this light is discarded. So only the light that is in focus will reach the detector. This will create a high signal-to-noise image.

So, again, we have the laser, goes into the beam splitter, is reflected to the objective, several rays. We have the in-focus light that goes through the objective and reach the detector, but we also have the out-of-focus light that also go through the objective but do not reach the detector because the pinhole will act as a blocker and will not allow the light to reach the detector. If this was an epifluorescence microscope, all the light would reach the detector and the image would be more hazy.

So as a summary, in a conventional light microscope, imagine we are imaging this block of beads and this is actually the focus area. In a conventional light microscope, this would be what I actually see in the field of view. In a confocal laser scanning microscope or a spinning disk microscope, this is the depth of focus, and this is the image in the field of view. I have removed all the beads that are out of focus and just have the image of the in-focus beads.

This is why in this image, you can see the widefield image with a lot of haze, the out-of-focus light from the planes above and below the plane of focus. And on the confocal image, you see a sharp image because the out-of-focus light was rejected by the pinhole.

This is the light source, the excitation goes into the objective and there is this Drosophila embryo. So all the light from all the planes of the Drosophila embryo will be captured by the detector, but only this orange is, in fact, in focus. In the confocal microscope, there will come light from all the planes also, from the in-focus plane and from the out-of-focus plane, but only the in-focus plane will reach the detector. The detector in laser scanning, point scanning microscope is actually different from the detector in a widefield microscope and it's PMT. There is no need to use a camera since the scanning is then point by point by point so there is no need to scan all the sample at the same time. Simultaneously, these detectors need to be fast because we are going to scan the sample point by point by point. So very briefly how they work, the incoming photon goes into the cathodes that will convert it into the electron. Once it's converted into the electron, it will reach the electrodes. The electrons will reach several electrodes and when they reach the electrodes, they will be able to jump out other electrons amplifying the signal a lot from a single photon up to a million times. The signal can now be read and the laser can move to the next point on the sample.

In point scanner microscopes, the pinhole can be adjusted to the objective, or even in a single objective, we can adjust the size of the pinhole and this is quite a big advantage of a confocal point scanning microscope. But what does it mean in practical terms to adjust, to increase or decrease the size of the pinhole? If I increase the opening of the pinhole, how does it impact on image intensity and on resolution? If I increase the opening of the pinhole, I will allow more light to go into the sample and, therefore, I increase the intensity. By increasing the pinhole, the Z section from the confocality will also increase, so more light from the planes that are out of focus will also reach the detector and, therefore, I will decrease resolution. On the other hand, if I close the pinhole, less light will reach the detector so I will decrease the intensity. But also by decreasing the pinhole, I'm increasing confocality and I'm diminishing the focal plane that I'm allowing to reach the detector and, therefore, I will increase the resolution.

So, as a summary, point scanners work by scanning line by line by line up until to form an image with the galvanomic mirror, moving it through the dichroic, through the PMT to amplify the signal and to form the image. Although the PMTs are extremely fast to detect a signal in a point in an instant, still the point scanners need to scan line by line by line up until forming an image. And for this reason, scan speed is a limiting factor in point scanning confocals.

So as a point scanner confocal summary, what happens to form the image is that the laser beam excites a point in the specimen. It will also inadvertently excite other points in the specimen but only the in-focus light is allowed to be detected by the photomultiplier tube because the out-of-focus light is rejected by the pinhole. The light detected by the PMT is associated to a picture on the monitor as you're seeing in this image here, and the laser beam will then move to the next point, and another pixel will be collected.

Again, although the PMTs are very fast, this will turn point scanning confocal imaging slower than widefield image.

The question would be, if there would be...if there is another way to acquire a confocal imaging without having to scan point by point an image? And the answer is yes, and it is done with spinning disk confocal microscopes. So there are advantages and disadvantages on both systems, both in spinning disk confocals and in point scanning confocals. I'm going now to present spinning disk confocal microscopes.

As I showed you before, point scanner work by collecting the light of a single pinhole. On the other hand, the spinning disk microscopes, which I will now introduce, will collect the light of multiple pinholes.

So as an example of a spinning disk, I'm showing you the dual microlens system. So the laser comes, and there's microlens that spins. The spinning disk is named like this because the disk that has the pinholes, it's spinning very fast, and the speeds will depend on the manufacturer, but there is a collector disk in the dual microlens system that will allow the light to be focused, the laser light to be focused on the pinholes. And both the collector disk and the pinhole disk, they must spin very fast but synchronically in a way that the lens...the light that is focused by the lens will reach the pinhole. Multiple pinholes are scanned simultaneously in a spinning disk, and the image, as the disk is spinning, the image is formed in an instant. And because the disk is spinning, multiple pinholes are scanned simultaneously. In the spinning disk, the detector is a camera.

So as comparing spinning disk confocal with point scanning confocal, the spinning disk, the disk is spinning, and the image is formed quite fast, whereas in the point scanning confocal, there's no disk spinning, there is a pinhole, and there's a laser that goes line by line by line to form the image. In the spinning disk, the detectors are cameras sCMOS or EMCCDs. On the point scanners, the detectors are the PMTs as I previously described. If you want to know a bit more on cameras, you have the previous courses or this course that camera technology and detectors was explained. So cameras are actually more efficient in collecting the photons that are emitted from the sample up to 90% quantum efficiency, whereas the PMTs have lower efficiency than the cameras.

So what are the challenges in spinning disk confocals. In fact, the signal-to-noise ratio of a sample of an image specimen is high...is higher if the specimen has a low background. But if the background is high, we will definitely need a confocal, but there are challenges in spinning disk, and the biggest challenge in the spinning disk is the problem that because the disk have multiple pinholes, if the pinholes do not have the appropriate size or spacing, we could have pinholes crosstalk, and we would lose confocality, we will acquire out-of-focus light from the neighboring pinholes. And this actually on traditional spinning disk confocals, this actually...pinhole crosstalk will actually start to appear in samples that are over 30 micron thick. They were not that useful for thicker samples.

And this is the importance of pinhole size and spacing in the spinning disk confocal microscopes. In traditional spinning disks as I was just talking in the previous slide, we have the in-focus light going through the pinhole but because the spacing was not appropriate or the size, the out-of-focus light was also captured, so there was no way to go very deep into the sample without losing confocality or capturing more background light and, therefore, decreasing the signal-to-noise ratio of the image we are acquiring. On new generation spinning disks, the pinholes' size and spacing have been optimized so actually modern spinning disks can image much deeper than 30 microns, for hundreds of microns without pinhole crosstalk.

On this slide, I want to present to you the importance of pinhole size and spacing in spinning disk confocal microscopes. The pinhole pitch is the space between the pinholes, and the pinhole diameter is obviously the diameter of the pinhole. And in this graph, I present you with comparing the rejection of background light with the thickness of a sample from 10 micrometer thick up to 80 micrometer thick. So, as for example, for the black line, we have a 50-micrometer pinhole with 253-micrometer pitch which can reject some background light in a 10-micrometer sample, but more than half of the light, background light, is captured in the image. It doesn't reject much of the background light in an 80-micron sample, meaning that this spacing does not allow it to go very deep inside the samples. With a 100-micron pinhole, 580-micron pitch, we have more rejection of the background light either in 10-micrometer samples or even in 80-micrometer samples. And this will increase with different designs, as for example, with a 50-micrometer pinhole and 580-micron pitch or with a 40-micron pinhole and custom pitch. We have actually a very high percentage of rejection of the background light even in an 80-micrometer sample.

And these are actually examples of spinning disk confocal images. This is a Drosophila embryo, and the point that I want to show you here is that actually spinning disk confocal images allow you the acquisition of quite fast images so this was nearly 300 images in 2 channels, and this was acquired in 77 seconds. And this image is 300 microns depth. So we can actually image deep inside cells and tissues with spinning, with modern technology spinning disks.

Another example to show that we can image deep inside cells and tissues with modern technology spinning disks is this cleared. This is a cleared tissue perfused with red beads, 8,000 optical sections in the 20x objective but just to show, this is 2.4-millimeter depth.

And this is another example of spinning disk imaging, the Drosophila egg chamber acquired in the spinning disk confocal image where you can image deep inside the samples as I previously showed.

Also, before I forget to mention, spinning disks were brought into the market in the early 2000s because they are actually an alternative for quite fast live imaging rejecting the out-of-focus light. Spinning disk allow very gentle imaging and fast imaging and imaging for a very long time, and this is another example of spinning disk confocal imaging in mouse tissues and analyzing in the intestines the deep imaging of the blood flow at 200 frames per second.

I've been talking to you about spinning disks and point scanners but which microscope should you choose for your specific application?

The first thing I want to tell you is that you can use any microscope that you have in your facility or in your lab. You can try and optimize the settings for the specific microscopes. These are actually...what I'm presenting you are guidelines that if you have more than one microscope to choose, maybe it's better to try this way or that way, but if you don't, just try with a microscope that you have. You might need to tweak and to work hard to get your data, but I believe in many cases you might get there. For a fixed thin image, widefield is always a good option. It's fast, it's cheap, so it's a good option. For a large, multi-tile imaging because the spinning disk is also quite fast, the spinning disk is a very good option. For thicker samples, both the spinning disks and the point scanners are much more appropriate than a widefield. And for more thicker samples, generally, the point scanners overtake the spinning disks. On a large, multi-tile image, the point scanner would also be useful but because if it's a very large image, it might take a longer time to acquire, then the spinning disk could be a better option because you save time. It's much faster to get your data. On the other hand, if you want to use spectral unmixing, if use fluorochromes in which the spectrums overlap then you'll definitely need a point scanner, but if you are using very low light imaging, the choice would come into a spinning disk because the detectors are much more sensitive.

Then as for a live sample considering the temporal resolution, and now I'm talking about a full field of view, acquiring the field of view...maximum field of view that you can get. For a thin sample, the widefield and the spinning disk are actually the best option either for low temporal resolution or high temporal resolution. For a thicker sample, most often I would choose a spinning disk, but the point scanners as with the low temporal resolution, might also work depending if you want to...of the temporal resolution you want or if you might want...if you need to go faster, you might also need to crop the field of view and, therefore, you can also use the point scanners, although for faster acquisitions, the spinning disk with thicker samples is...it would be always the best option.

And this is an example of a neuronal cell, the image in spinning disk confocal microscope in a Dragonfly confocal microscope.

And this is another example of a spinning disk confocal microscope imaging delivering fast imaging at three frames per second.

And with this, I finish my talk. I thank you very much for listening to this talk, and I will be happy to take any questions and answer to you later.