Detectors for Live-Cell Imaging

Date:2019/7/12 13:35:37

Modern imaging of living cells in culture relies heavily on electronic area array detectors, including CCD cameras, as well as high-gain point source detectors, such as photomultiplier tubes (PMTs). Although differing in architectural detail, both the digital camera and the photomultiplier are analog sensors built on a common principle. Photons impinging on a photoactive substrate are converted to photoelectrons and ultimately read out as an analog current or voltage stream. Digitizing the signal, by serially approximating the number of electrons in the stream at specific times, creates a relative gauge of light intensity, free from the progressive accumulation of noise inherent in storing analog data. By correlating the sampling time with the position of the array element or the scanned beam, the intensity values are spatially correlated to recreate an image of the specimen. Detector settings that are user-definable typically include exposure time, readout rate, gain, offset, and selecting a region of interest (a portion of the array) for readout.

In all electronic detectors, signal generation occurs when photons strike the active substrate material on the detector surface. The exposure time and illumination intensity determine the number of photons generated for image formation. The efficiency of converting photons to photoelectrons is governed by the fraction of the detector surface that can sense photons (known as the fill factor in CCDs) combined with the quantum efficiency of the substrate. The quantum efficiency is defined as the number of photons successfully converted divided by the total number impacting the substrate. In addition, the detector material inevitably generates a number of spurious electrons, ominously termed dark current. Cooling the image sensor, often to a range between -20 and -80 degrees Celsius, suppresses the dark current when long exposure times are necessary.

After being collected by the detector, the signal and dark current electrons are read out to an analog-to-digital converter (ADC). Reading the electron output results in an artifact known as readout noise, which occurs either from thermally generated electrons or from small timing errors, and is also directly proportional to readout speed. Together, the dark current and readout noise constitute the principle noise sources in the image recording process, beyond the statistical nature of photon counting (referred to as shot noise). The driver software for many digital cameras permits selection of a readout amplifier speed, usually specified in megahertz (MHz) or pixels per second. Choosing faster readout frame rates results in lower signal-to-noise ratios. Similarly, in confocal microscopy, the beam scanning speed relative to the sampling time of the photomultiplier, can often be altered to increase the signal being read out to the converter.

Three basic variations of CCD architecture are in common use for scientific imaging systems: full frame, frame transfer, and interline transfer (illustrated in Figure 5). The full-frame image sensor (Figure 5(a)) has the advantage that nearly 100-percent of its surface is photosensitive, although it must be protected from incident light during readout (usually with an electromechanical shutter). Frame-transfer CCDs (Figure 5(b)) operate faster than the full-frame devices because exposure and readout can occur simultaneously. Although they are similar to their full-frame counterparts in structure, one-half of the rectangular pixel array is covered with an opaque mask that is used for a storage buffer for photoelectrons prior to readout. In contrast, the interline-transfer CCD (Figure 5(c)) contains columns of active imaging pixels along with masked storage-transfer pixels that alternate over the entire array. Because a charge-transfer channel is located immediately adjacent to each photosensitive pixel column, stored charge must only be shifted a short distance, rendering these chips the fastest in terms of readout. A majority of the color and cooled monochrome scientific digital cameras are equipped with interline-transfer image sensors, however, more advanced electron-multiplying CCDs (discussed below) typically contain frame-transfer chips.


The analog-to-digital converter samples the sensor readout over a tightly controlled timer interval, which is synchronized with the readout amplifier. This stage is where the photoelectrons and noise electrons from the detector are converted to a gray level or intensity value. The converter has an input range that limits the total number of electrons before saturation (white), and a set number of electrons that are required for creating each gray level. The term bit depth refers to the binary range of possible grayscale values employed by the converter to translate analog image information into discrete digital values capable of being read and analyzed by a computer. For example, the most popular 8-bit digitizing converters have a binary range of 2?(E8) or 256 possible values, while a 10-bit converter has a range of 2?(E10) (1,024 values), and a 16-bit converter has 2?(E16), or 65,536 possible values. The bit depth of the analog-to-digital converter determines the size of the gray scale increments, with higher bit depths corresponding to a greater range of useful image information available from the camera. Another important concept, the dynamic range is typically specified as the maximum achievable signal divided by the camera noise, where the signal is determined by the full-well capacity (maximum of photoelectrons that each pixel is capable of holding), and the noise is the sum of dark and readout noises. As the dynamic range of an image sensor is increased, the ability to quantitatively measure the dimmest intensities is improved.

A detector with 4 electrons of system noise probably uses at least four electrons for each gray level and would require an input range of 16,384 (4?(E12)) electrons to be digitized to 12-bit resolution. Note that for some image sensors (especially CCDs) with 12-bit (4096 gray levels), 14-bit (16,384 gray levels), or even 16-bit (65,500 gray levels) converters, the image detector material may not have sufficient full-well capacity to fill the entire analog-to-digital converter input range under all circumstances. In this case, more gray levels are generated by the analog converter than correspond to actual photoelectrons obtained from the detector. This is often the situation with commercial monochrome camera systems that digitize image data above 10-bits. The importance of having the high converter bit depths lies in the ability to capture real differences in signal intensity (in effect, contrast), even if the effort of doing so requires a smaller number of gray levels than are available to the sensor.

The signal from the digital camera or photomultiplier can be matched to the input range of the analog-to-digital converter by carefully setting the gain and offset parameters, as well as the readout amplifier speed. For example, if two adjacent pixel elements that have collected 101 and 109 electrons, respectively, are fed into an 8-bit converter with an input range of 2560 electrons, then 10 electrons would be required for each gray level. With no electronic gain, the adjacent pixels would have identical values (a grayscale level of 10) in the final image. If the combined detector noise for this system is only 2 electrons, the counts of 101 and 109 should be distinguishable. Applying a gain of 5 (a multiplicative process) yields 505 and 545 electrons, respectively, now sampled to values of 50 and 54 by the converter. Thus, the gain amplifier has increased the difference in the detector value such that it can now be interpreted by the analog converter as separate values. Because, in this hypothetical case, a gain factor of 5 would push any detector value greater than approximately 500 out of the converter input range, the offset control can be used to subtract electrons prior to digitization. Subtracting 400 electrons in the example above leaves 105 and 145 electrons, sampled to values of 10 and 14 in the resulting image. In practice, the gain amplifier introduces noise in proportion to the gain factor, limiting the effective multiplier.

In practice, the detector and illumination aperture should be adjusted to gather light from the smallest possible area of the specimen that still contains the information required for feature identification and context. Imaging a larger area increases specimen exposure to illumination and lengthens the sensor readout time. For digital cameras, the readout area can be set using the accompanying software package that controls acquisition. Matching the resolution of the objective to the image sensor resolution should also be taken into account. Laser scanning confocal microscopes also enable software-selectable areas for image acquisition, cleverly disguised as a zoom function. When using this feature, be aware of the interplay between the zoom factor and the spatial sampling rate, which is often recalculated by the microscope software. In cases where selecting a small area to image does not change the number of samples (pixels) proportionately, then the goal of shortening the illumination and readout time has not been achieved. In this situation, the image size should be reset with the sampling rate, bearing in mind the resolution limits of the objective.

About Nuohai Life Science

Nuohai Life Science co., Limited is committed to equipping scientists in life science field with advanced experimental and research & manufactural equipments, related consumables, reagents as well as offering professional application and technical support services. We aim to constantly promote the development of new technology in life science field with the timely introduction of the latest foreign technology and products. Meanwhile, Nuohai Life Science provide CRO and CMO service to acdamic and industrial customers, to satisfy every demand in the R&D and manufacturing. 

Nuohai Life Science not only brings new top products to our customers, but also pays attention to training professional sales, engineers and after-sales service team. With customer-oriented value in heart, Nuohai Life Science has earned the trust of our customers and enjoyed good reputation with comprehensive life science solution and sound after-sales service.

Nuohai Life Science was established in 2004 with headquarter in Shanghai and 8 offices in Beijing, Guangzhou, Chengdu and other first-tier cities, providing a nationwide selling & service networking. Customer service is just a call away.


Add:Room 1203, Block B, No. 650 Xinzhuan Road, Songjiang District, Shanghai, China



©Copyright 2017 锘海生命科学 沪ICP备17032315号