Biomedical Engineering Reference
In-Depth Information
Fig. 4.18 The principle of a computed tomography imaging spectrometer (CTIS). a The system is
constructed of a collimating lens, a two-dimensional holographic grating, (shown in b ), a focusing
lens, and an array detector (Image courtesy of Greg Bearman, Snapshot Spectra, USA)
a relatively small sample is analyzed covering a small portion of the array detector.
If that is the case, there are practically a lot of extra pixels on the detector that are
not being used, especially with the growing number of pixels in newer version of
array detectors.
This is exactly what the method of computed tomography imaging spectrometer
(CTIS) does, by taking advantage of a clever two-dimensional dispersion element.
The method, sometimes termed as snapshot spectral imaging, should be distin-
guished from the methods previously described in Sect. 4.4.1 by the fact that the
latter ones enable a very limited number of wavelengths to be measured while CTIS
can provide a high-resolution spectral image.
CTIS allows to measure a full spectral image in a single exposure time [ 46 , 47 ].
It is based on a holographic phase plate that acts as a two-dimensional dispersion
element that projects the dispersed spectral-spatial information to the array detector.
The spectral image is reconstructed from the raw data by using adequate algorithms,
and the size of the FOV is traded off for the spectral information [ 48 ].
The schematic description of the system is shown in Fig. 4.18 a. The real image is
collimated with a lens and the collimated light passes through a two-dimensional
dispersion element whose structure is shown in Fig. 4.18 b. After passing this
element, the light is imaged with another lens onto the detector.
The outcome of CTIS is a typical pattern shown in Fig. 4.19 .Theimageis
constructed of 5 5 sub-images. The central one shows the non-dispersed image
and does not contain spectral information. The other images are the ˙ 1 and
˙ 2 diffraction orders of the image along the two main array axes, with a radial
dispersion. By using different holographic dispersion elements, other patterns can be
generated, for example, with smaller number of diffraction orders thereby increasing
the size of the FOV [ 46 ]. This data is analyzed by respective types of deconvolution
algorithms that take into account the transfer function of the system. The transfer
function is measured during the calibration of the system and includes a series of
images that each of it describes the pattern that is measured on the detector when
illuminating a point at a given wavelength. This can be achieved, as an example, by
Search WWH ::




Custom Search