optimization of the detector placement [22]. The optimization is based on
minimizing the Compton scatter X-rays, which are the primary source of interference in X-ray fluorescence imaging of low quantities of contrast agent. The study
proposed an optimized anisotropic spatial orientation of detectors such that there is a
low probability of acquiring scatter X-rays which have the same energy as fluorescence X-rays. In this optimization, X-ray detectors are placed at large scattering
angles relative to the primary excitation beam. At large scattering angles, the
Compton scatter X-rays loose enough energy in the scattering process to be separable from the fluorescent signal.
As a consequence, this optimized detector configuration allows higher molecular
sensitivities which was validated in Monte Carlo simulations. A mouse-sized phantom, with a diameter of 2.25 cm, containing various low gold concentrations
(10–100 μg/ml) of nanoparticles expected in molecular imaging applications was
excited with different beam spectra (monoenergetic, Gaussian, and tungsten anode).
It is shown that the optimized configuration, in which X-ray detectors cover only a
specific portion of a sphere surrounding the imaging volume, performed better than
the isotropic configuration (in which detectors cover the entire sphere). This was a
counterintuitive result that forces a redesign of XF imaging instrumentation. The
optimization results in a tenfold sensitivity improvement over isotropic detection for
gold nanoparticles and established a theoretical sensitivity limit of 10 μg/ml. The
sensitivity improvement significantly enhanced the potential of the X-ray fluorescence modality for future preclinical and clinical molecular imaging.
7.2 Compton Camera XFCT
The anisotropic detection principle, described above, can also be combined with
Compton imaging [50]. The position and energy information of a fluorescent X-ray
can be collected with a Compton camera. This information is used to restrict the
possible emission of XF X-rays to a cone-surface in the image domain. Here we talk
about electronic collimation on the detector side in comparison of physical collimation on the source side in pencil-beam XFCT. In contrast to physical collimation on
the detector side, the electronic collimation reaches higher sensitivity as no valuable
XF photons are absorbed by a collimator.
With the spatial information from the Compton camera, scatter events can be
distinguished based on their emission direction which is specifically efficient for
deep tissue imaging. This enables reasonable molecular sensitivity in clinical in vivo
XF imaging and allows high-speed imaging as the target region can be excited with
fan- or cone beams. Image reconstructions in our Monte Carlo studies were able to
recover 5 mm diameter lesions in the center of a human lung. The results show that
low concentrations down to 150 μg/ml can be detected for a CdTe energy resolution
of 110 eV at 78 keV and a target dose of 30 mGy. Since the detector developments
are not yet advanced enough to realize XFI with the Compton Imaging concept,
further advances in detector instrumentation are required to realize clinically feasible
Compton camera XF imaging.
X-Ray Excited Fluorescent Materials for Medical Application
139
minimizing the Compton scatter X-rays, which are the primary source of interference in X-ray fluorescence imaging of low quantities of contrast agent. The study
proposed an optimized anisotropic spatial orientation of detectors such that there is a
low probability of acquiring scatter X-rays which have the same energy as fluorescence X-rays. In this optimization, X-ray detectors are placed at large scattering
angles relative to the primary excitation beam. At large scattering angles, the
Compton scatter X-rays loose enough energy in the scattering process to be separable from the fluorescent signal.
As a consequence, this optimized detector configuration allows higher molecular
sensitivities which was validated in Monte Carlo simulations. A mouse-sized phantom, with a diameter of 2.25 cm, containing various low gold concentrations
(10–100 μg/ml) of nanoparticles expected in molecular imaging applications was
excited with different beam spectra (monoenergetic, Gaussian, and tungsten anode).
It is shown that the optimized configuration, in which X-ray detectors cover only a
specific portion of a sphere surrounding the imaging volume, performed better than
the isotropic configuration (in which detectors cover the entire sphere). This was a
counterintuitive result that forces a redesign of XF imaging instrumentation. The
optimization results in a tenfold sensitivity improvement over isotropic detection for
gold nanoparticles and established a theoretical sensitivity limit of 10 μg/ml. The
sensitivity improvement significantly enhanced the potential of the X-ray fluorescence modality for future preclinical and clinical molecular imaging.
7.2 Compton Camera XFCT
The anisotropic detection principle, described above, can also be combined with
Compton imaging [50]. The position and energy information of a fluorescent X-ray
can be collected with a Compton camera. This information is used to restrict the
possible emission of XF X-rays to a cone-surface in the image domain. Here we talk
about electronic collimation on the detector side in comparison of physical collimation on the source side in pencil-beam XFCT. In contrast to physical collimation on
the detector side, the electronic collimation reaches higher sensitivity as no valuable
XF photons are absorbed by a collimator.
With the spatial information from the Compton camera, scatter events can be
distinguished based on their emission direction which is specifically efficient for
deep tissue imaging. This enables reasonable molecular sensitivity in clinical in vivo
XF imaging and allows high-speed imaging as the target region can be excited with
fan- or cone beams. Image reconstructions in our Monte Carlo studies were able to
recover 5 mm diameter lesions in the center of a human lung. The results show that
low concentrations down to 150 μg/ml can be detected for a CdTe energy resolution
of 110 eV at 78 keV and a target dose of 30 mGy. Since the detector developments
are not yet advanced enough to realize XFI with the Compton Imaging concept,
further advances in detector instrumentation are required to realize clinically feasible
Compton camera XF imaging.
X-Ray Excited Fluorescent Materials for Medical Application
139