Multimodal Biomedical Imaging Systems - Biomedical Optical Imaging Technologies: Design and Applications - page 322

9.3.2.5 Detectors

The detector is one of the key components in biomedical imaging because it
determines at what level the signal can be detected, what relevant structures can be
resolved, and the dynamics of the process that is visualized and recorded. Through
the detector, the optical signal is converted into electrical energy, thereby producing
a measurable signal that is proportional to the number of photons detected. The
measured, continuous analog signal is then converted to a discrete digital signal.
Given that there are a limited number of intensity conversion levels, 8, 10, 12 or
16 bit, it is unavoidable that this conversion process introduces a certain amount
of error.
Three commonly used light detectors are photodiode, PMT, and solid-state
imaging sensors. The first two employ a photosensitive surface that detects incident
photons and generates an integrated electronic signal, but these devices lack spatial
discrimination. PMTs can be used to detect the incoming photons at high speed
because they do not store a charge and can respond to changes of the input light
within a few nanoseconds. PMTs typically generate a low noise signal and have a
large dynamic range to accurately reflect the photon flux. PMTs also have large gain
without sacrificing bandwidth.
Silicon photodiode and avalanche photodiode devices (APDs) also respond
very quickly to incoming light but without large gain. Compared to PMTs, the
spectral sensitivity of photodiodes is relatively flat and the quantum efficiency
(QE) is relatively high over the entire visible spectrum. The major drawback with
photodiodes is that they produce a considerable amount of noise, resulting in a low
SNR under photon-limited conditions, which is the case for fluorescence imaging.
Another disadvantage, compared to PMTs, is that the photosensitive surface is
smaller than the PMT.
Solid-state imaging sensors consist of a dense matrix of photodiodes incorpo-
rating charge storage regions. CCD and CMOS detectors are two commonly used
solid-state detectors in biomedical imaging. Solid-state sensors can be characterized
by a number of parameters including QE, dynamic range, uniformity, SNR, response
speed, and spatial resolution.
The CCD is the most widely used imaging sensor in biomedical imaging. While
a CCD sensor is usually operated at room temperature, the performance can be
significantly enhanced by cooling the sensor to reduce the dark current. Although
the CCD consists of silicon photodiodes, its spectral sensitivity is different from
that of a simple silicon photodiode detector because the CCD surface has channels
used for charge transfer, which are shielded by polysilicon gate electrodes. These
structures absorb the shorter wavelengths, resulting in a lower blue sensitivity of
the device. These losses can be eliminated in the back-illuminated CCD where light
falls onto the back of the CCD in a thinned and transparent region.
There are a number of new CCDs developed for some special applications,
such as slow scan CCD, intensified CCD (ICCD), and electron multiplying charge
coupled devices (EMCCD). Slow scan CCDs are designed specifically for low