Multimodal Diffuse Optical Imaging - Biomedical Optical Imaging Technologies: Design and Applications - page 353

pair, in transmittance configuration, was raster scanned over the compressed breast
area. However, the implementation of reconstruction methods was crucial to provide
information that was clinically useful [ 30 , 31 ].
A tomographic system based on six laser diodes operating at wavelengths from
660 to 836 nm modulated at 100 MHz has been developed at Dartmouth College
[ 32 ]. Light delivery and collection is performed via fiber bundles that are rigidly
coupled to a 2D circular ring which rotates such that one of the 16 fiber bundles in
the measured plane is aligned with the light source coupler while the other 15 fiber
bundles are each connected to a photomultiplier tube (PMT) detector. The fibers
are directly in contact with the pendulant breast, and light pressure is applied to
ensure proper contact. The system does not require matching liquid. Acquisition is
performed in the same plane, but multiple planes can be investigated for 3D imaging.
FD instrumentation is also used to supplement CW techniques. A hybrid
CW frequency domain device for optical tomography has been developed at the
University of Pennsylvania [ 33 ]. The system combines four amplitude-modulated
laser diodes (690, 750, 786, and 830 nm) which are rapidly switched between 45
optical fibers on a 9 5 array and detected in reflectance, whereas transmitted
light is detected by a CW CCD camera. The information obtained by the FD
channels supplement the spatially dense CW information for enhanced functional
tomography [ 34 ].

10.3.3 Time Domain

The first implementations of time-resolved techniques for diffuse optical imaging
were motivated by the notion that minimally scattered photons (ballistic photons)
enabled the rendering of images similarly to X-ray computed tomography (CT) with
a relatively good resolution [ 35 ]. However, such photons can be reliably detected
only for a few millimeter-thick samples, and the idea was abandoned. Time domain
refers to techniques based on a source that produces a short pulse of light (typically
less than 200 ps) and detectors recording the time of flight of photons exiting
the sampled volume with a temporal resolution in the tenth of picoseconds or
less. The temporal distribution of photons detected after propagation is known as
the temporal point spread function (TPSF). Such data set is the richest in terms
of information content and provides an unmatched set of data type: amplitude,
mean time of flight, variance, skew, and Laplace transform [ 36 ]. Such data types
provide enhanced separation between absorption and scattering contributions over
the frequency domain. However, such benefits come at the cost of more expensive
instrumentation and greater difficulty to implement in clinical prototypes. Also,
TD optical imaging is a photon-starved technique that requires relatively long
integration time (a few seconds compared to subsecond for CW and FD systems) to
obtain TPSF with good statistical distributions.