Multimedia Library

WEBINAR: Optimizing Live Cell Confocal Imaging

What's the best drift compensation precision for both lateral and axial?

We have most experience of the ZDC from Olympus and PFS from Nikon with little use of DF from Zeiss so far. To date the PFS has proven to be fastest (20ms) and therefore most effective for live cell work. PFS can be used in "continuous" mode, while in our hands ZDC is used for single shot application, but is still effective. One thing to be careful of is whether a particular objective is supported for autofocus control – some are simply "not compatible".

With only 1/2 the camera pixels, how do you handle low mag, ie 20X?

The answer is with half the field of view. At 20X we can still achieve excellent background rejection, but with reduced confocality and hence better SNR. Even at 10X there is good performance with smaller disk patterns e.g. 80 µm. At 10X we would have twice the field of view of the 20X objective. So we expect there will be cases where the selection of a different disk pattern will be advantageous.

Can you show us some movies of live cell imaging with the DSD confocal?

We have some movies, but they have been captured with the 80 µm disk pattern and this is really the thing that made us realize we needed to redesign the disk. So right now we cannot show high quality live cell movies. However, in July we have an imaging fest planned and will publish those movies as soon as we can do so. The optimization for live cell imaging demands the trade off between confocality and sensitivity. This has now been resolved and is really the message behind the webinar.

With regard to axial chromatic aberration i.e. from uv to red, which lenses do you recommend for multifluorescence experiments using the DSD?

This is an interesting question and depends a great deal on the microscope manufacturer that you use. In our experience you will only get the advertised performance from selected objectives. Back in 1992 I published a paper with well known colleagues, to characterize chromatic aberration in about forty different microscope objectives form the major microscope companies. We used the tandem scanning confocal microscope of Petran and Hadravsky’s design. Only one objective met the specification. Things have been improved since then, but tolerances still mean that optimal performance is rarely met. Olympus offers, for additional cost, objectives selected for PSF and chromatic performance, but this very fact indicates that tolerances require hand picking. Generally Plan-apo-chromats should be corrected at 3 or 4 wavelengths and so they should show best performance. But to be sure, characterise your objectives – measure PSF at the wavelengths of interest and then choose. When working in the UV it will always be challenging to remain parfocal across all wavelengths.

Reference Akinyemi, O., Boyde, A., Browne M.A., Hadravsky, M., Petran, M.“Chromatism and Confocality in Confocal Micorscopes” Scanning, Vol 14, pp136-143 1992.

How straining is the system to the specimen? Do you need 2 x the excitation light hitting the specimen for the same amount of signal?

You say twice as much light...implying 50% transmission...well it is 50% transmission, but remember when we combine the signal we are effectively getting the benefit of twice confocal over the widefield so it’s difficult to answer and I think we would have to dig into what we are comparing it to. If you would like to add more detail, I will be happy to engage in this question further.

Can you vary the speed of the disk?

It could be done but there is currently no reason to do this. The disk runs at 6000 RPM which means that one revolution takes 10 ms. In our experience 10 ms is the minimum exposure at which you will obtain sufficient signal to form and image. Increments of 10 ms are convenient and since 100 Hz exceeds the target imaging speed of the instrument we don’t see any point in doing so. Further, we are synchronized to the disk if we run our exposure at multiples of 10 milliseconds but we have found this is not critical when you exceed one revolution of the disk. This contrasts with the CSU, where you achieve one scan in 30 degrees of the disk, which translates to 1 ms at 5000 rpm. So in the CSU we increment our exposure by 1ms increments for disk synchronisation.

Can you summarise which applications the spinning disk can compete with as standard? i.e. is there any specific application suited to the spinning disk?

• It is targeted at live cells and whole embryos.
• For embryos we can work at low power and still achieve good confocality, so we can reject out of focus light without cross-talk problems, which is a problem with other spinning disks.
• For live cells we can tune the optical sectioning to get plenty of signal in a relatively short time.
• So compared to point scanners it will win in terms of imaging speed. While compared to a laser spinning disk e.g. Yokogawa CSU it performs much better with thicker specimens, but cannot compete for pure speed. Hence, the DSD will find niche applications based on flexible illumination wavelength, moderate speed, low background and cross-talk in thick specimens and cost-effectiveness.

How do you change, in the system, between low and high magnification modes? Is this motorized?

Switching between the modes is the click of a button in the software interface. The disk is motorized and slides the relevant part of the disk and hence light pattern into the light path. It takes about 2 seconds to do switch between these patterns.

When do you expect to have the solution for the upright microscopes?

The upright solution will be completed in the July/August time frame and will ship in November or December.

What microscopes do the Photonics Inst, products fit on currently?

Pretty much every microscope on the market.

Will it be possible to have multi-dimensional real-time reconstructions in Imaris, with the data being acquired on an Andor system?

• The two things on our roadmap to integrate between Imaris and iQ are as follow.
  
  • Allow Imaris to read the ImageDisk directly which should be available in July.
  • Real time rendering is actually an extension of the ability for Imaris to peak in at the images as they being acquired and this is on the roadmap for December of this year.