X-ray imaging performance of scintillator-filled silicon pore arrays
Cross-sectional view of a pore array after electrochemical etching (tilt: ) before filling with CsI(Tl). The pitch is , the walls are , and the depth is about .
Thickness maps for two samples (top No. 1, bottom No. 6). The inset in the bottom image gives the CsI thickness in . Note, that these maps are based on x-ray absorption measurements that do not resolve every single pore. The circular area defines the etched and CsI-filled region. The dimensions of the images are .
Recalculated CsI thickness histograms for all samples (from maps as shown in Fig. 2). Note, that the thickness is derived from x-ray absorption measurements that do not resolve every single pore. The sample Nos. 1–4 show a similar mean value and width of the distribution.
SEM image of a cross section from sample No. 3. The voids at the bottom of some columns appear to stem from the filling procedure rather than from the cross-section preparation. The polycrystalline nature of the solidified CsI is also visible.
Schematic setup for x-ray measurements, not to scale.
Sensitivity of three samples as histograms, normalized to their maximum values.
Signal vs dose for sample No. 3 for two different beam qualities in a double logarithmic plot. The results for the slope are from a regression calculation.
Experimental results of the MTF measurement with the edge method for three samples. For comparison the MTF of the aperture (sinc function) and the MTF of commercial columnar CsI, measured on the same sensor, is also shown.
NPS of sample No. 3 vs spatial frequency, plotted for various dose levels and two different beam qualities. In addition, the dark NPS, i.e., without x ray, is shown.
DQE vs spatial frequency for sample No. 1 for various applied doses and two different beam qualities ( and ).
DQE vs spatial frequency for the best sample (No. 3).
Comparison of the DQE of two pore array samples measured with beam quality at a dose of with the DQE of a commercial columnar CsI:Tl sheet at a dose of .
DQE at a fixed spatial frequency (0.5 lp/mm) vs dose for sample Nos. 1, 2, and 3.
X-ray images of line pair phantoms with 7.5 lp/mm (top) and 9 lp/mm (bottom). The black parts on the right and lower left are outside the scintillator-filled area. The black line in the center is caused by two adjacent defect columns of the sensor, which are not interpolated by the defect correction algorithm.
X-ray image of the fingertip of an anatomical hand phantom. The black areas are outside of the scintillator-filled region.
Combination of two x-ray images showing a joint in the little finger of an anatomical hand phantom.
Light guiding efficiency and optical Swank factor as a function of mean scatter length calculated from simulation results for three different beam qualities.
Top: x-ray absorption of sample No. 3 (, ), measured with a flat x-ray detector (not resolving each pore). Bottom: Corresponding gain map of the pore array in combination with the high-resolution optical CMOS sensor ( signal, signal). Hardly any correlation between the two images is seen.
Top: x-ray absorption of sample No. 6 (, ). Bottom: Corresponding gain map ( signal, signal). Many of the black spots can be found in both images, representing low CsI filling and at the same time very low light output as indicated by the white ellipse.
Theoretical contribution of pixel aperture, fluorescence and optical gap to the total MTF compared to the measurement results of sample No. 3.
Data on the beam qualities used for most of the measurements. is the thickness of Al placed in the beam. Together with the internal filtration an equivalent thickness of is achieved. is the number of quanta per area and dose as calculated by the simulation program.
Data on the three beam qualities that were explicitly verified for their defining half value layers. is the tube voltage for which the right HVL is reached experimentally.
Definition and numbers for the quantities used in Eq. (4). The absorption is dependent on the beam quality and is here shown for as example.
Results for the light guiding efficiency of the three best samples.
DQE(0) for different samples and for two beam qualities with estimated errors. is the mean x-ray quantum energy corresponding to the used beam quality.
DQE(0) for the best sample (No. 3) for different beam qualities.
Factors contributing to the DQE. Theoretical and experimental results for the DQE(0) of three samples.
Theoretical and experimental results for one sample (No. 3) and three different beam qualities.
Article metrics loading...
Full text loading...