A guide to fluorescence imaging cameras

Sensitivity and signal-to-noise ratio

Sensitivity describes a camera's ability to detect weak light signals.

A highly sensitive camera produces a strong signal relative to background noise, creating a high signal-to-noise ratio (SNR). Improving signal-to-noise ratio can be achieved by increasing the signal collected, reducing electronic noise, or increasing exposure time.

Because colour cameras use colour filter arrays that block a proportion of incoming photons, they are typically less sensitive than equivalent monochrome cameras. For this reason, monochrome cameras are generally preferred for fluorescence imaging unless simultaneous multi-colour image acquisition is required.

High sensitivity is particularly important when imaging:

• Low-light fluorescence signals
• Calcium transients
• Live-cell preparations
• Fast cellular events
• Single-molecule fluorescence
• Electrophysiology experiments
• Small changes in fluorescence intensity

As imaging systems have evolved, improvements in sensor sensitivity have enabled researchers to acquire higher-quality data while using lower illumination intensities, reducing photobleaching and phototoxicity.

Quantum efficiency (QE)

Quantum Efficiency (QE) measures the proportion of incoming photons successfully converted into electrons. For example:

• A camera with 50% QE converts approximately one in every two photons into signal.
• A camera with 90% QE converts approximately nine out of every ten photons into signal.

Higher QE generally improves image quality, particularly when signal levels are low. In fluorescence imaging, where emitted signals are often weak, high QE allows more of the available photons to contribute to image formation.

QE is especially important when imaging:

• Low-light fluorescence
• Calcium signals
• Live-cell preparations
• Time-lapse experiments
• Single-molecule fluorescence

When illumination levels are high, even a camera with relatively modest QE may still collect enough photons to generate a strong image. However, under low-light conditions, higher QE can have a significant impact on image quality and signal-to-noise ratio.

Manufacturers often specify a peak QE, which represents the wavelength at which the sensor converts photons to electrons most efficiently. Since QE varies with wavelength, researchers should consider QE performance across the wavelengths relevant to their fluorophores, rather than focusing solely on the peak value.

Modern scientific CMOS (sCMOS) sensors routinely achieve QE values above 70%, while premium back-illuminated (BSI) sensors can exceed 90%.

Read noise

Read noise refers to electronic noise generated as the sensor converts collected charge into a digital signal. Read noise effectively limits the smallest fluorescence signal that can be detected.

Low read noise is essential for:

• Dim fluorescence
• Fast imaging
• Short exposure times
• Weak biological signals

Modern scientific cameras have dramatically reduced read noise compared with previous generations, allowing researchers to detect faint signals that would previously have been lost in the noise floor.

Dark noise and sensor cooling

Dark noise is generated by thermal activity within the sensor itself.

Even in complete darkness, some electrons are produced purely by heat, which can reduce image quality during long exposures. As the dark noise is dependent on the exposure time this type of noise is especially important for applications where a long exposure time is required, such as in astronomy. The implication of dark noise for biological applications is generally limited, especially for electrophysiology where a short exposure time is often preferred.

Most scientific cameras employ cooling systems to minimise dark noise and improve image clarity.

Common cooling methods include:

• Thermoelectric cooling

• Forced-air cooling

• Water-assisted cooling

• Fanless cooling designs

For electrophysiology applications, minimising vibration and electrical interference can be much more important than minimising dark noise. Fanless camera designs can therefore offer significant advantages by eliminating potential sources of vibration and electrical noise.

Frame rate

Frame rate describes the number of images captured per second (fps). Fast frame rates are critical for studying dynamic biological processes such as:

• Calcium signalling
• Voltage imaging
• Neuronal activity
• Membrane dynamics
• Cell migration
• Fast electrophysiological events

Several factors influence frame rate:

• Sensor architecture
• Sensor size
• Exposure time
• Region of Interest (ROI)
• Data transfer interface
• Computer performance

Maximum frame-rate specifications can sometimes appear impressive, but practical performance is ultimately limited by the amount of available light. Capturing thousands of frames per second is only useful if sufficient photons are reaching the sensor to generate meaningful image data.

Pixel binning

Binning combines neighbouring pixels into a larger “super pixel”. In 2 × 2 binning, four adjacent pixels are combined and read out as a single pixel.

Benefits include:

• Increased sensitivity
• Improved signal-to-noise ratio
• Faster image acquisition

The trade-off however is reduced spatial resolution.

Binning remains a valuable technique for low-light imaging where sensitivity is more important than maximum image detail.

Modern fluorescence imaging relies increasingly on advanced sCMOS technology, which combines high sensitivity, low noise, large fields of view and fast image acquisition.

While CCD and EMCCD technologies continue to serve specialised applications, most researchers now benefit from the performance advantages offered by modern sCMOS cameras.

By understanding key parameters such as quantum efficiency, read noise, dynamic range and frame rate, researchers can make informed decisions and select the camera best suited to their imaging workflow.

Whether performing patch clamp electrophysiology, calcium imaging, live-cell imaging or fluorescence microscopy, choosing the right camera remains one of the most important investments in achieving high-quality experimental data.