CMOS will be an affordable alternative to sCMOS sensors in biomedical applications

As technology advances, metal-oxide-semiconductor (CMOS) technology for complementary medicine already has the advanced imaging capabilities required for many biomedical applications, but can it replace expensive sCMOS (scientific CMOS) sensors? CMOS and sCMOS sensors have established benchmarks for the performance and value of machine vision in several industries. Today Xiaofei will explain the advantages and costs of these two technologies in extremely demanding biomedical and life science imaging applications.

The difference between CMOS and sCMOS

Generally, sCMOS sensors are considered “next-generation” CMOS sensors.

sCMOS technology was introduced to bridge the gap between new CMOS sensors and traditional CCD (Charge Coupled Device) sensors early in CMOS development. Initially, CMOS sensors were not available for biomedical applications due to limitations in dynamic range, read noise, frame rate, and resolution.

When sCMOS cameras were introduced, they used very similar design principles and fabrication techniques to CMOS sensors, but combined several features to help them overcome the initial shortcomings of CMOS. This makes sCMOS sensors suitable for scientific applications requiring low-light performance, wide dynamic range and high fidelity.

However, since the introduction of sCMOS cameras, conventional CMOS sensors have gained huge improvements in quantum efficiency and reducing their own internal noise, making CMOS cameras a viable option for many advanced biomedical applications. And most CMOS cameras are much more affordable than sCMOS cameras. This factor alone motivates many engineers and researchers to consider the latest CMOS sensors when choosing a microscope camera, histology camera, cytology/cytogenetics camera or epifluorescence camera for their application.

Products Recommended

If you choose to use a CMOS camera, you can choose the FLIR Backfly S series, which is more commonly used for epi-fluorescence applications.

The Blackfly S camera series offers a very rich sensor and interface (both USB3 and GigE). Sensor options in boxed and board-level form factors are also extensive.

The choice of CMOS and sCMOS

Choosing a CMOS or sCMOS sensor depends on a number of factors. If you’re torn between the two, you can use epi-fluorescence because white light is sufficient even without buying an sCMOS sensor. Whether one is more appropriate than the other is often just a matter of the amount of light reaching the camera, or a combination of specific performance parameters for a specific application.

Whether it’s CMOS or sCMOS, you should choose a monochromatic sensor over a color sensor because of the inherent quantum efficiency advantage of monochromatic sensors.

sCMOS sensors feature backside illumination and larger pixels (eg, CCD technology) that reduce overall noise. Additionally, sCMOS cameras often include a Peltier cooling system to reduce thermally generated noise after long exposures. Cameras using sCMOS sensors also require a high-bandwidth interface, such as CameraLink or CoaXpress with a frame grabber board, making such vision systems more complex and therefore more expensive.

To address this problem, CMOS manufacturers are continually improving quantum efficiency (the ability to collect incident photons), reducing read noise (ensuring a lower degree of incident photons are not lost in the noise) and employing backside illumination. While some CMOS sensors can also use Peltier cooling systems, improvements in quantum efficiency and noise reduction have made cooling unnecessary for some biomedical imaging applications.

Another way to reduce costs is through interfaces. For years, CMOS sensors have been paired with consumer interfaces such as USB3, GigE, and 10 GigE. These interfaces do not require frame grabbers, reducing system complexity (and thus cost). The upcoming 25/100GigE, USB4, and CXPX interfaces will help to completely solve this problem with significantly increased bandwidth.

Products Recommended

For CMOS cameras, you can also choose the FLIR Oryx series.

The Oryx camera family has high-resolution sensors compatible with the fastest 10GigE interface. Oryx cameras are full-featured and suitable for taller end applications, but in a larger form factor. Oryx is a good choice if transfer speed is important.

CMOS is an attractive alternative to sCMOS

The lower cost alone has prompted many engineers and system designers to consider evaluating the latest CMOS sensors to replace sCMOS-based systems. In many cases, vision system designers are pleasantly surprised to find a suitable CMOS camera for less than $1,000, whereas a typical sCMOS setup with similar performance parameters might cost $10,000.

Whether sCMOS or CMOS, many camera manufacturers do not use a single standard to compare cameras. So comparing cameras can be a challenge no matter what type of sensor is used. In the field of machine vision, EMVA1288 has become a recognized camera specification and measurement standard by the European and American AIA (American Automatic Imaging Association) and Japan’s JIIA (Japan Industrial Imaging Association).

To sum up, for situations where extreme performance is required, an sCMOS camera may be a must. However, it is necessary to determine the most important performance parameters for your specific application, and to make a reasonable comparison of CMOS and sCMOS cameras before eliminating each other.

CMOS sensors are constantly evolving, and the price gap between CMOS and sCMOS is rapidly closing. If a conventional CMOS sensor can meet your application requirements, this may be a much cheaper alternative for you and your team.

Product choice

Both the FLIR Blackfly S and Oryx camera series can be controlled and programmed with GenICam3 and the Spinnaker SDK for faster application creation. To further narrow your camera model selection, use the FLIR Machine Vision Model Selector, which includes 14+ (EMVA 1288 based) imaging parameter filters. To find models that perform well in low light conditions, filter for higher values ​​for absolute sensitivity, quantum efficiency, and dynamic range.