The spinning disk confocal microscope (SDCM) is a revolutionary tool for imaging in the life sciences, observing samples ranging from single molecules to live cells, featuring high speed, 3D and multichannel acquisitions. Many experiments and researchers use SDCM imaging systems for their imaging, and the technology has become well established. This article will explore the current applications that SDCMs are being used for, the capabilities of modern systems, and how SDCMs are likely to advance and evolve in the future.
Applications of Spinning Disk Confocal Microscopy
SDCM allows for optical sectioning (imaging 2D planes from thick 3D samples) at high speed with considerably reduced photodamage or bleaching when compared to widefield. When paired with a highly sensitive camera, SDCM becomes a powerful technique for imaging in the life sciences.
SDCM is primarily used for multichannel, high speed, high-resolution 3D imaging of live samples, potentially over the long term. Thicker samples can be used to allow for 3D imaging, such as zebrafish and brain tissues, but smaller, low-light samples can benefit from SDCM, such as neuronal circuits, actin cytoskeletons, and mitochondria.
Evolving Spinning Disk Confocal
SDCMs can be improved in a number of ways, from the type of detector scientific camera to the format and speed of the disk. See how your spinning disk imaging can evolve:
- Avoid EMCCDs or front-illuminated sCMOS cameras
- Use more modern back-illuminated sCMOS camera technology
- Combine large field of view microscopes and sCMOS cameras for more efficient imaging
- Advance to super-resolution levels with an extra set of microlenses
- Combine high disk rotation speeds with a high-speed sCMOS camera in order to capture more from a live dynamic sample
- Use multiple sCMOS cameras to capture multiple wavelengths simultaneously
This article will explore each of these points in more detail, including other applications for SDCM.
Avoid EMCCD, Use Back-Illuminated sCMOS Cameras
Previous iterations of confocal microscopes used laser scanning (LSCM), a single pinhole and alternates to cameras, including photomultiplier tubes (PMTs) or avalanche photodiodes (APDs). These alternates featured low quantum efficiency (QE), which is the rate at which photons of light are converted to electrons of charge in order to produce an image. With low QEs, information is lost due to light scattering or inefficient sensor designs. The QE of PMT/APD ranged from 20‑40%, meaning that these imaging systems operated at suboptimal efficiencies and limited the signal to noise ratio.
Far higher QEs are attainable with cameras, around 70-80% for front-illuminated sCMOS, and ~90% with EMCCDs. However, some of the best QEs available from modern cameras are found in back‑illuminated sCMOS, which features 95% QE at the peak, such as the Prime and Kinetix families from Photometrics. This near-perfect QE allows imaging systems to maximize efficiency and boast higher sensitivities.
EMCCD vs Back-Illuminated sCMOS
In addition, by using back-illuminated sCMOS researchers can avoid complications associated with EMCCD cameras. EMCCDs have a limited lifespan due to gain decay, where higher EM gain values will decay over time, meaning they require regular calibration for any quantitative measurements or reproducible experiments. In addition, temperature affects gain stability, meaning EMCCDs often include aggressive cooling and have a very large form factor. Excess noise factor is also present, where EM multiplication will also multiply up other sources of noise such as photon shot noise, reducing the signal to noise ratio. On top of these drawbacks, EMCCDs are still one of the most expensive formats of scientific cameras, even now 20 years after their introduction. By switching to back-illuminated sCMOS cameras, researchers can both enjoy the benefits of larger sensors, higher speeds, and higher QEs, as well as avoiding the numerous pitfalls associated with EMCCDs.
For a direct comparison between an EMCCD and the Prime 95B on a spinning disk system, see our tech note and Fig.1.
The advantages of the Prime 95B over EMCCD are also outlined by Dr. Peter March of the University of Manchester, who is using SDCM for long-term, high-speed 3D imaging of live cells, as seen in Fig.2. See the full customer story on our website.
Front-illuminated sCMOS vs Back-illuminated sCMOS
Front-illuminated sCMOS cameras feature a lower QE but can also use split sensors, this often results in artifacts and fixed pattern noise, all of which affect the acquired image at low light. The effect of a front-illuminated split sensor vs a back-illuminated Prime 95B single sensor can be seen in Fig.2.
The use of a modern back-illuminated sCMOS allows for the highest currently-available peak QE, a clean bias lacking patterns/artifacts, and huge improvements in field of view (FOV) and speed over EMCCD. Future SDCM imaging systems will most likely make use of modern and next-gen back‑illuminated sCMOS cameras, such as the Prime 95B, Prime BSI or Kinetix.
Increase Efficiency With Larger Fields Of View
Older models of SDCMs were designed with a field of illumination that would match to typical CCDs or EMCCDs, with approx. ~12 mm diagonal field of view (FOV). Pinhole diameter, spacing, microlens focal length and light source illumination was also designed to best match these factors accordingly. However, this meant that older SDCM models were limited in terms of their FOV, especially when microscope FOV can range from 18-25 mm. In this case, camera technologies and SDCMs were compromising the overall FOV of the imaging system, as seen from Fig.3.
Modern sCMOS cameras have far larger sensors and a much greater FOV. The Prime 95B back-illuminated sCMOS camera has both 22 mm and 25 mm FOV models, and the new Kinetix sCMOS has an enormous 29.4 mm FOV, this is approx. six times more camera area than older CCD or EMCCD models, resulting in far more data with every frame. Some FOV options from the Photometrics range of cameras can be seen in Fig.4, compared to CCD and EMCCD.
With large FOV sCMOS cameras, far more (or all) of the information from the microscope can be captured and recorded. This presents the opportunity for large FOV spinning disk systems that can capture more information with every frame. With large FOV systems, fewer images are needed to capture information from the sample, meaning experiments can be more streamlined, and photosensitive samples are less exposed to light and can be maintained for longer. Less stitching and processing is needed to construct large XY or 3D images, and experiments are generally simplified.
Modern SDCM systems feature larger microlenses with longer focal lengths and larger dichroic mirrors, resulting in FOV up to 25 mm. This makes these systems best compatible with modern sCMOS cameras, greatly increasing the available FOV for SDCM.
It should be noted that large FOV SDCM systems also require suitable illumination, as seen in Fig.3 some systems and higher light intensity in the center of the frame than around the edges, with a larger FOV these issues can be exacerbated. Improvements in FOV for spinning disk have been accompanied by improvements in illumination, such as multi-mode fiber illumination and more powerful lasers.
Image Even Smaller Samples With Super-Resolution Spinning Disk
As outlined in our article on super-resolution spinning disk confocal microscopy, there are spinning disk systems that can break the diffraction limit of light and image samples below ~200 nm, while retaining the high speed and 3D optical sectioning of SDCM.
Essentially, due to the dual pinhole setup of confocal microscopes, there are two point spread functions (PSF) in a SDCM: the excitation PSF and the emission PSF. These two PSFs are offset, by narrowing and altering these PSFs the resolution of the system can theoretically improve by √2x, up to 2x after deconvolution.
This narrowing of PSFs could be done with far smaller pinholes, but this blocks too much light. Instead, optical photon reassignment is performed using another set of lenses on the underside of the pinhole disk, narrowing the PSF and reassigning photons to the most probable position they originated from, as seen in Fig.5.
These super-resolution SDCM systems can have a maximum resolution limit of ~120 nm (with standard SDCM it is around ~250 nm) and can image live samples at up to 200 fps (by rotating the disk at 4000 rpm) using most fluorescent dyes. This opens up many options for both researchers doing super-resolution imaging looking for more speed and better 3D imaging, as well as established SDCM researchers who are looking to resolve smaller samples.
The super-resolution (and single-molecule localization microscopy (SMLM)) techniques photoactivated localization microscopy (PALM) and stochastic optical reconstruction microscopy (STORM), are also compatible with SDCM. To find out more about PALM and STORM, please refer to our article on super-resolution localization microscopy. These techniques have also been shown to benefit from super-resolution processing algorithms such as super-resolution optical fluctuation imaging (SOFI).
Cameras For Super-Resolution Spinning Disk
Because these systems have such a high maximum resolution, it is necessary to pair these systems with a modern camera that is capable of sampling at these super-resolution levels, featuring high sensitivity and speed. Back-illuminated sCMOS cameras offer an ideal option for these systems, but it is important to pair the camera with the appropriate optics depending on pixel size. For example, the Prime 95B sCMOS features an 11 μm pixel, while the Prime BSI, BSI express, and Kinetix sCMOS cameras feature a smaller 6.5 μm pixel. These cameras could be used with a 240x or 280x magnification respectively, as this allows for Nyquist sampling at 62 nm and 90 nm, either of which is suitable for oversampling from 120 nm.
The smaller pixel size of a 6.5 μm camera may give slightly improved resolution but at the cost of reduced sensitivity, detecting 2.2x fewer photons than an 11 μm pixel camera. On the other hand, an 11 μm pixel camera improves contrast and allows for exposure times to be reduced while still oversampling enough for effective deconvolution. This highlights the importance of pixel size when selecting which back-illuminated sCMOS to pair with a super-resolution SDCM system.
The ability to image at higher resolutions allows researchers to reveal smaller samples, as well as unveiling details about existing samples. By upgrading existing SDCM systems to super-resolution levels, powerful images of live biological samples can be taken with high sensitivity, super-resolution, and high speed. This technique is capable of reliable super-resolution observations in biomedical routine research and is a vital part of advanced microscopy.
Dr. Joana Delgado Martins of the University of Zurich is using a super-resolution SDCM system with two Prime BSI cameras to image a variety of samples, including time-lapse imaging of live zebrafish embryos as seen in Fig.6. See the full customer story on our website.
Capture More By Increasing Disk Rotation Speed
With a typical SDCM and Nipkow disk, each area of the image is scanned by a single pinhole with approx. every 30° rotation of the disk. By rotating the disk faster, the maximum image acquisition speed can be increased.
For example, a disk with 50 μm pinholes spaced 250 μm apart will achieve 12 frames per revolution, resulting in a potential 1000 frames per second (fps) if spun at 5000 revolutions per minute (rpm). If this disk is rotated faster, the acquisition speed will increase.
But not all SDCM systems have similar disks. Another system with a larger FOV requires larger microlenses in order to increase the focal length. These larger microlenses require larger pinholes or greater pinhole spacing. A large FOV SDCM system with 50 μm pinholes and 500 μm pinhole spacing only achieves 3 frames per revolution, meaning that the speed is far lower, around 200 fps at 4000 rpm. In this way, different SDCM systems can achieve different speeds due to their disk patterns. But for all these systems, the faster the disk can be spun, the higher the maximum acquisition speed.
At the time of writing, spinning disk systems could achieve speeds up to 15,000 rpm, resulting in very high potential acquisition framerates. However, acquisition at these high speeds can only occur when a high-speed SDCM system is paired with a high-speed camera.
Cameras For High-Speed Spinning Disk
When capturing high-speed dynamic processes such as calcium imaging, short camera exposure times are necessary. These exposure times need to be as close as possible to the time taken for the disk to scan the image, in order to best capture high-speed processes. If exposure time less than the scan time, part of the image is not scanned and black lines streaks will appear on the image, as seen in Fig.7.
The solution to this problem at high acquisition speed is to synchronize the camera exposure to the scanning of the disk, ensuring that the exposure time is equal to a whole number multiple of the scan time. This requires both a high-speed camera and access to advanced triggering options, so that camera acquisition can be synchronized to the disk rotation. Photometrics sCMOS cameras feature both, with high speeds and access to hardware triggering.
The Kinetix sCMOS can achieve speeds up to 500 fps across the full 29.4 mm sensor, by cropping this sensor the speed increases linearly, easily achieving several thousand fps and matching to even the fastest SDCM systems. In addition, the Kinetix can also achieve over 100 fps in the high sensitivity low read noise CMS mode, allowing for a combination of high speed, large FOV and high sensitivity. A high speed back-illuminated sCMOS camera allows for the full potential of SDCM systems to be utilized, especially if the disks can rotate at speeds of up to 15,000 rpm.
Simultaneous Multichannel Imaging With Multi-Camera Systems
In a SDCM, each camera can be paired with a filter wheel in order to perform multichannel imaging. With multiple filter wheels, multiple cameras can be used in order to simultaneously image two different wavelengths. This avoids the need to manually or electronically switch the filter wheel to a different wavelength, which is a slow process (seconds to milliseconds) and can cause information to be lost, especially if the aim is to image multiple fast dynamic processes.
Many modern SDCM systems feature the ability to connect multiple cameras for two-channel detection, future systems may support up to four different cameras for simultaneous multichannel detection, or other multi-camera applications such as multi-focal imaging or imaging at different polarizations. In addition, certain techniques such as Förster Resonance Energy Transfer (FRET) can benefit from multiple cameras.
The multiple cameras on a SDCM can be the same model of camera, for experimental consistency, or markedly different cameras. Cameras with different specifications could be used on a single SDCM system, allowing for the benefits of different FOVs, pixel sizes and magnifications on the same sample. By switching between cameras and objectives, vastly different experiments can be done on the same SDCM system with relative ease.
Additional Future Advances For Spinning Disk Systems
Modifiable Pinhole Disks
As certain disk pinhole sizes are optimized for different camera pixel sizes and objective magnifications, the ability to change pinhole size on the fly would be advantageous. The smaller the pinhole, the greater the maximum resolution of the system, but the more light is blocked. Combining smaller pinholes with low magnification, high numerical aperture lenses and a highly sensitive back‑illuminated sCMOS camera would allow for SDCM imaging across a large FOV without sacrificing resolution. Opening up the pinholes would then allow for more illumination and better contrast, at the cost of pinhole cross-talk, but these could be paired with optimal fluorescent probes to lessen the effect.
By having a SDCM system where disks can either be swapped out or a single disk can be modified, the ability to change both the pinhole diameter and pinhole spacing would allow for a great degree of experimental flexibility. By having two different sCMOS cameras on this system would also allow for optimal imaging at both ends of the pixel size/magnification spectrum.
While SDCM can produce high-resolution images, especially with super-resolution SDCM, this can be further increased by computational post-processing steps such as deconvolution. These steps can be time-consuming and computationally intensive, but more options are emerging for real-time deconvolution, both in the SDCM system and on board modern sCMOS cameras. Several open‑source 3D deconvolution options are available for standard computers to run, allowing researchers to increase the maximum resolution of their system.
Multiphoton Spinning Disk
As described in our two-photon microscopy application note, multiphoton imaging uses two or more near infrared (NIR) photons in order to image deep (up to several millimeters) into a 3D sample due to different absorption coefficients between visible light and IR. By pairing multiphoton imaging with SDCM, high speed imaging can be performed deep in living tissue, presenting a powerful platform when working with large model organisms.
However, it should be noted that multiphoton SDCM relies upon multiple pulsed infrared lasers across multiple pinholes, limiting the illumination area.
By making changes to spinning disk pinholes, microlenses, and rotation speed, imaging with a SDCM can be flexible and change to your experimental needs. However, one of the best ways to evolve your SDCM imaging is using a modern back-illuminated sCMOS camera, which will benefit your imaging speed, resolution, sensitivity, and field of view, adapting to your sample and imaging requirements.
To find out more, refer to our Applications page on spinning disk confocal microscopy to find out the best cameras for this application, and how they can be demoed on your system.
Azuma, T. and Kei, T. (2017) Development of high-speed super-resolution confocal scanner. Yokogawa technical report English Edition. Vol 60(2) 33-37.
Oreopoulos, J., Berman, R., & Browne, M. (2014). Spinning-disk confocal microscopy. Quantitative Imaging in Cell Biology, 153–175. doi:10.1016/b978-0-12-420138-5.00009-4
Stehbens, S., Pemble, H., Murrow, L., & Wittmann, T. (2012). Imaging intracellular protein dynamics by spinning disk confocal microscopy. Methods in enzymology, 504, 293–313. https://doi.org/10.1016/B978-0-12-391857-4.00015-X