A novel approach to accurate portal dosimetry using CCD-camera based EPIDs
Configuration of the additional lead collimator surrounding the pencil beam at the most off-axis position without (A) and with (B) the lead plug in the collimator opening. This collimator was used in the PB-acquisition protocol (see text).
Example of horizontal cross-talk kernel interpolation method, applied to 1D data. “Experimental” kernel (thick solid line), at induction position (i.e., center positions of the applied PB) (indicated with vertical thin solid line) and kernel (thick broken line), at induction position (vertical thin broken line). The shifted kernels (thin solid line) and (thin broken line) are combined into the derived kernel (thick dashed-dotted line) for induction position via weighted averaging (see text). For illustration, the difference in peak value between the two experimental kernels has been exaggerated. The open circles indicate the region of kernel , that is filled-in using kernel data, by applying the kernel symmetry characteristics.
EPIs for an on-axis PB, with the intensity scale normalized to the CEPID signal maximum. The EPI was obtained using the PB acquisition protocol (see text), and is displayed at full intensity range (A) and with limited intensity range, without (B) and with the convolution processing applied (C), white solid lines indicate iso-intensity contours). (D) EPI acquired using the standard acquisition protocol (see text) (E) Cross sections in the direction at of the EPIs obtained using the PB acquisition protocol (thick line) and the standard acquisition protocol (thin line) [i.e., of the EPIs displayed in Figs. (C) and (D), respectively].
Cross sections in the direction at of EPIs for an on-axis PB position, (A) acquired using the PB acquisition protocol at four days in an period, (B) acquired using a position of the additional collimator on the treatment couch of (thin broken line), (i.e., as used in the PB acquisition protocol) (thick solid line) and below the target (thin solid line), and for a position of below the target but with a miss-positioned collimator opening (circles), (C) acquired using a collimator opening width of (i.e., as used in the PB acquisition protocol) (solid line) and (broken line).
(A) EPI for on-axis PB position obtained using the PB acquisition protocol, after replacing the data of a circular region around the PB position (edges indicated with dotted line) by a fit of the data outside this region using a normal distribution. (B) EPI signal of the center pixel row (; dots) and the fit using a normal distribution (solid line). The vertical broken lines indicate the edges of the interpolated region.
CEPID signals at the position of the slits in the mask of the CEPID fluorescent screen, for PB positions at and , 0, and (upper, center, and lower panel, respectively). The EPIs are obtained using the PB-acquisition protocol for three configurations described in the methods and materials section: The standard CEPID (thick solid line), CEPID with masked fluorescent screen (thin solid line) and for a masked fluorescent screen with an extra lead collimator between treatment couch and CEPID (thin broken line). Open circles in lower panel: Fit with inverse square law relation (see text). Inset in each panel: CEPID fluorescent screen (large gray square), with the position of the PB (black square) and the slits in the fluorescent screen mask (white rectangles on both sides of the PB position).
Overview of the cross-talk kernels in the calibration set of PB positions at and at a grid. The white squares indicate the PB position. Each panel shows the cross-talk distribution for the complete CEPID fluorescent screen, with and ranging from at isocenter plane. The white lines indicate intensity contours (levels 0.001, 0.002, …, 0.005).
(A) Measured cross-talk kernel for PB position. (PB position indicated with filled white square). (B) Kernel for same PB position as (A), but obtained via the two-step interpolation scheme from the measured 30-position cross-talk kernel calibration set (PB positions indicated with open squares). (C) Difference of experimental minus “two-step interpolated” cross-talk kernel.
Ratio of the on-axis CEPID signal over the on-axis dose , measured with the LDA with build-up, for on-axis square fields of (open circles, thin broken line), (open triangles, thin solid line), (filled circles, thin broken line) and (filled triangles, thick solid line), as a function of the thickness of the homogeneous polystyrene phantom. For each field width, the ratios are normalized to the mean value measured for that field width.
CEPID local sensitivity calibration matrix .
Results for cranial and caudal off-axis Alderson-phantom thorax fields, the field shapes are indicated in the overview image (A) (and also in the other panels) using solid and broken lines, respectively. (C) and (D) show the EPI-based PDIs (converted to the accumulated diode charge dose scale), (E) and (F) the cross-talk contributions and (G) and (H) the EPI-based minus LDA-based PDI relative differences for the cranial and the caudal off-axis field, respectively. The -method settings were and 1% for the maximum tolerable DTA and dose difference , respectively. In (G) and (H), the dotted line indicates the position of the LDA reference diode with mount, present while scanning the LDA. (B) Cross sections in the -direction of the EPI-based PDI (solid line), LDA-based PDI (filled circles), cross-talk contribution (dashed-dotted line) and the EPI-based PDI before cross-talk correction (dashed line).
Results for step-and-shoot IMRT field: (A) EPI of this field, with contours indicated for gray-levels 200 and 230, (B) EPI-based PDI, with contours indicated for accumulated diode charges of 600 and , (C) EPI-based minus LDA-based PDI relative difference, for -method settings of and 1% for the maximum tolerable DTA and dose difference , respectively. (D) Cross sections in the -direction of the EPI-based PDI (solid line), LDA-based PDI (filled circles), cross-talk contribution (dashed-dotted line) and the EPI-based PDI before cross-talk correction (dashed line).
EPI-based minus LDA-based PDI on-axis difference (in %) for a series of square on-axis fields for three thicknesses of a homogeneous polystyrene phantom and for two asymmetrical fields for polystyrene phantom thickness. Also the on-axis cross-talk fraction is indicated (in % of the on-axis EPI signal) for the applied square fields in case of phantom thickness.
Article metrics loading...
Full text loading...