If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Takasaki Institute of Advanced Quantum Science, Foundational Quantum Technology Research Directorate, National Institutes for Quantum Science and Technology (QST), Japan
Takasaki Institute of Advanced Quantum Science, Foundational Quantum Technology Research Directorate, National Institutes for Quantum Science and Technology (QST), Japan
Takasaki Institute of Advanced Quantum Science, Foundational Quantum Technology Research Directorate, National Institutes for Quantum Science and Technology (QST), Japan
Prompt X-ray imaging during irradiation with spread-out Brag peak (SOBP) beams of carbon ions.
•
Prompt X-ray imaging of SOBP beams was performed with an MLC.
•
Prompt X-ray images of SOBP beam shapes at clinical dose levels could be obtained with the X-ray camera.
•
The SOBP images obtained in this experiment serve as evidence that our prompt X-ray imaging system can be used in clinical conditions.
Abstract
Prompt secondary electron bremsstrahlung X-ray (prompt X-ray) imaging using a low-energy X-ray camera is a promising method for observing a beam shape from outside the subject. However, such imaging has so far been conducted only for pencil beams without a multi-leaf collimator (MLC). The use of spread-out Bragg peak (SOBP) with an MLC may increase the scattered prompt gamma photons and decrease the contrast of the images of prompt X-rays. Consequently, we performed prompt X-ray imaging of SOBP beams formed with an MLC. This imaging was carried out in list mode during irradiation of SOBP beams to a water phantom. An X-ray camera with a 1.5-mm diameter as well as 4-mm-diameter pinhole collimators was used for the imaging. List mode data were sorted to obtain the SOBP beam images as well as energy spectra and time count rate curves. Due to the high background counts from the scattered prompt gamma photons penetrating the tungsten shield of the X-ray camera, the SOBP beam shapes were difficult to observe with a 1.5-mm-diameter pinhole collimator. With the 4-mm-diameter pinhole collimators, images of SOBP beam shapes at clinical dose levels could be obtained with the X-ray camera. The use of a 4-mm-diameter pinhole collimator attached to the X-ray camera is effective for prompt X-ray imaging with high sensitivity and low background counts. This approach makes it possible to image SOBP beams with an MLC when the counts are low and the background levels are high.
In particle ion therapy, it is desirable to obtain the beam range or shape measurements during beam irradiation from outside the subject, since a particle ion beam forms a Bragg peak at the end of the range. The range of this therapy is mainly determined by the particle ions’ energy and the materials that cross their paths [
]. Therefore, when anatomical variations occur between the acquisitions of the planning computed tomography, the path length of the particle ion therapy changes and results in reduced target coverage or increased dose to the organs at risk [
]. Therefore, on-line monitoring of the beams is desired in particle ion therapy.
Several methods have attempted to image the beam from outside the subject. Prompt gamma imaging using a slit camera or Compton cameras has been attempted in efforts to estimate the beam [
]. Imaging of positrons produced in subjects by irradiation with particle ions is another method used for imaging the beam. Imaging of the produced positrons has been performed using a positron emission tomography (PET) system, a high-energy gamma camera, and a Compton camera [
Patient study on in-vivo verification of beam delivery and range using positron emission tomography and computed tomography imaging after proton therapy.
Imaging of the prompt secondary electron bremsstrahlung X-rays (prompt X-rays) emitted from a subject during particle-ion irradiation is also a promising method for beam range or shape imaging [
M. Yamaguchi, Torikai K, Kawachi N, Shimada H, Satoh T, Y. Nagao Y, et al., Beam range estimation by measuring bremsstrahlung, Phys Med Biol. 2012; 57,2843-56.
Development of a low-energy x-ray camera for the imaging of secondary electron bremsstrahlung x-ray emitted during proton irradiation for range estimation.
Yamaguchi M, Nagao Y, Ando K, Yamamoto S, Sakai M, R. Parajuli RK, et al., Imaging of monochromatic beams by measuring secondary electron bremsstrahlung for carbon-ion therapy using a pinhole x-ray camera, Phys Med Biol. 2018; 63(4),045016.
Development of a YAP(Ce) camera for the imaging of secondary electron bremsstrahlung x-ray emitted during carbon-ion irradiation toward the use of clinical conditions.
]. It employs a pinhole X-ray camera to image the beams during irradiation of subjects with particle ions. Imaging of the beams was successful for the pencil beams not only with uniform phantoms but also with non-uniform phantoms [
Up to now, prompt X-ray imaging during irradiation with particle ions has been conducted only for pencil beams without a multi-leaf collimator (MLC). Furthermore, no imaging experiment has been attempted for a spread-out Bragg peak (SOBP) with an MLC for particle ions. The SOBP imaging condition using an MLC may increase the scattered prompt gamma photons from the MLC while also decreasing the contrast of the images. Although spot-scanning therapy systems have become popular for protons without using an MLC, some particle-ion therapy systems still use a wobbler method combined with an MLC to form SOBPs. The use of that latter approach leads to the most serious condition for imaging a prompt phenomenon from outside the subject. Prompt X-ray imaging of the SOBP with an MLC is required before prompt X-ray imaging is used for clinical trials because SOBP with an MLC would produce high uniform background counts distributed in the field-of-view from scattered prompt gamma photons. We define this uniform distribution of counts in the field-of-view as background counts.
Consequently, in this paper we performed prompt X-ray imaging for SOBPs formed with MLCs during irradiation by carbon ions to a water phantom. To improve sensitivity and reduce the background fraction in the images, we tried using a larger-diameter pinhole collimator for the measurements.
2. Methods
(1) Imaging experiments.
The set-up used for the imaging experiments on carbon-ion SOBP beams is schematically shown in Fig. 1. A water phantom was set on the bed of the carbon-ion therapy system (Mitsubishi, Japan). On a spacer, we set a cerium-doped yttrium aluminum perovskite (YAP(Ce)) pinhole X-ray camera. The distance between the pinhole camera and the center of the phantom was 40 cm. We employed the same pinhole YAP(Ce) camera used previously in similar experiments [
Sensitivity improvement of YAP(Ce) cameras for imaging of secondary electron bremsstrahlung X-rays emitted during carbon-ion irradiation: Problem and solution.
We used two types of tungsten pinhole collimators, where one was 1.5 mm in diameter and the other was 4 mm in diameter. The 1.5-mm-diameter pinhole collimator was the same as that previously used for pencil beam imaging [
Sensitivity improvement of YAP(Ce) cameras for imaging of secondary electron bremsstrahlung X-rays emitted during carbon-ion irradiation: Problem and solution.
], while the 4-mm-diameter pinhole collimator was a newly developed device. One reason for using the 4-mm-diameter pinhole collimator was to improve sensitivity. Each pinhole collimator had a height of 30 mm with an angle of ∼ 50 degrees. The exterior of each collimator was cut into a screw shape to make it an exchangeable component for the pinhole X-ray camera. Theoretically, the sensitivity of a 4-mm pinhole collimator will improve seven times over that of a 1.5-mm collimator, with a sacrifice in spatial resolution. This is because the sensitivity of pinhole X-ray camera is proportional to square of the pinhole diameter. The expected spatial resolution at 40 cm from the camera was ∼ 30 mm FWHM and ∼ 50 mm FWHM for the collimators with 1.5-mm and 4-mm pinholes, respectively.
Another reason for using the larger-diameter pinhole collimator was to reduce the relative background noise level. Since the background counts in the prompt X-ray images were from the detection of the prompt gamma photons penetrating the tungsten shield, the background counts of the 4-mm pinhole collimator were roughly the same as those of the 1.5-mm collimator. With the increase in sensitivity from using the 4-mm pinhole collimator, the background fraction of the 4-mm pinhole collimator will decrease to ∼ 1/7 that of the 1.5-mm collimator.
(2) Beam shapes.
The experiment was performed at the Hyogo Ion Beam Medical Center (HIBMC) in Japan. We used carbon ions in this experiment because we could previously measure the prompt X-ray images for carbon-ion pencil beams in the facility [
Yamaguchi M, Nagao Y, Ando K, Yamamoto S, Sakai M, R. Parajuli RK, et al., Imaging of monochromatic beams by measuring secondary electron bremsstrahlung for carbon-ion therapy using a pinhole x-ray camera, Phys Med Biol. 2018; 63(4),045016.
Development of a YAP(Ce) camera for the imaging of secondary electron bremsstrahlung x-ray emitted during carbon-ion irradiation toward the use of clinical conditions.
Sensitivity improvement of YAP(Ce) cameras for imaging of secondary electron bremsstrahlung X-rays emitted during carbon-ion irradiation: Problem and solution.
]. An SOBP beam irradiated to the water phantom is schematically illustrated in Fig. 2. The SOBP beam was generated with a wobbler magnet and a scatterer by circulating the beam, a ridge filter was used to generate SOBP in the depth direction, and an MLC was used to form the lateral beam shapes. For our experiments, we irradiated SOBP beams made from 241.5 MeV/n carbon ions to a 20 cm × 20 cm × 10 cm water phantom for 90 s and acquired all of the count data using a data acquisition system in list mode. The intensity of the carbon-ion SOBP beams was ∼ 1 × 107 /s, ∼1/100 that of the pencil beam (∼1 × 109/s). The range of the SOBP beam without a range shifter was 10 cm, and the size of the SOBP area was 5 cm (X) × 5 cm (Y) × 5 cm (Z). We carried out the imaging without a range shifter (RS: 0 cm), with a 2-cm range shifter (RS: 2 cm), and with a 4-cm range shifter (RS: 4 cm) for the SOBP beams. The carbon-ion beam shape was rectangular, where the width and depth were 5 cm × 5 cm for conditions both with and without range shifters. Even with the range shifters, these dimensions did not change because the SOBP beams were cut and shaped to be 5 cm × 5 cm with the MLC after penetrating the range shifter.
Fig. 2Dimensions of SOBP beam for imaging of prompt X-rays during irradiation by carbon ions to water phantom.
Development of a low-energy x-ray camera for the imaging of secondary electron bremsstrahlung x-ray emitted during proton irradiation for range estimation.
Yamaguchi M, Nagao Y, Ando K, Yamamoto S, Sakai M, R. Parajuli RK, et al., Imaging of monochromatic beams by measuring secondary electron bremsstrahlung for carbon-ion therapy using a pinhole x-ray camera, Phys Med Biol. 2018; 63(4),045016.
Development of a YAP(Ce) camera for the imaging of secondary electron bremsstrahlung x-ray emitted during carbon-ion irradiation toward the use of clinical conditions.
]. The weight-summed analog signals were converted to digital signals in the data-acquisition system. The signals digitized by the A-D converter (100 MHz) above a threshold level were digitally integrated for 320 ns and used for position calculation based on the Anger principle. These digitally integrated signals were also used for energy information. The data-acquisition system was basically the same as that used for our PET systems [
Yamamoto S, Watabe H, Kanai Y, Watabe T, Kato K, J. Hatazawa J. Development of an ultrahigh resolution Si-PM based PET system for small animals. Phys Med Biol. 2013; 58,7875-88.
], with some modifications to achieve a smaller size. Digitized signals larger than the threshold level were also digitally integrated for 150 ns and employed to calculate the pulse-shape spectrum based on the dual integration method by calculating the ratio of a 150-ns integration value to a 320-ns integration value [
A time stamp was generated by reading the value of a 1-ms timer. The calculated position (X: 9 bit, Y: 9 bit), energy (7 bit), pulse shape (7 bit), and time stamp (16 bit) were fed to a laptop personal computer (PC) set outside the carbon-ion therapy room via a ∼ 40-m-long local area network (LAN) cable.
(4) Image processing.
The measured list mode image data were sorted into image data with desired pixel sizes (maximum: 512 × 512 pixels), energy widths (maximum: 128 channels), pulse shape widths (maximum: 128 channels), and temporal widths (minimum: 1 ms). The sorted images were stacked using public domain software (ImageJ) and corrected for non-uniformity by dividing the uniform image measured by an Am-241 gamma photon (60 keV) image using the same software. We also analyzed the energy spectra from the stack images using the same software. The sizes of the images were basically 512 × 512 pixels. The sizes of the images used to derive the energy spectra were 128 × 128 pixels.
Depths and lateral profiles for the measured images were set with the widths of 20 pixels (20 mm) for 512 × 512 pixel images. Ranges were determined by fitting the falling part of the depth profiles with linear approximation, and the depths at 10 % of the peaks were measured. We evaluated several areas for the fitting and derived the average and standard deviations (SDs) for the measured images. Each width of the beam was determined by Gaussian fitting the lateral distribution and full width at half maximum (FWHM), and SD was measured by software (Origin, 2018B). Time sequential images were also derived from list mode data with 100-ms and 1-ms intervals. For the time sequential images, changes in the time count rate were measured for the SOBP beam areas, and time count rate curves were calculated.
3. Results
3.1 Prompt X-ray images of SOBP beams with 1.5-mm-diameter collimators
Fig. 3(A), (B), and (C) show prompt X-ray images of a water phantom with a 1.5-mm-diameter pinhole collimator during irradiation of SOBP carbon-ion beams without a range shifter, with a 2-cm range shifter, and with a 4-cm range shifter, respectively. Only low-contrast SOBP images can be observed in the central part of the images.
Fig. 3Prompt X-ray images with a 1.5-mm-diameter pinhole collimator of SOBP carbon-ion beams without a range shifter (A), with a 2-cm range shifter (B), and with a 4-cm range shifter (C).
Fig. 4(A) and (B) show depth and lateral profiles, respectively, of prompt X-ray images of a water phantom with a 1.5-mm-diameter pinhole collimator during irradiation of SOBP carbon-ion beams. We could observe some shapes of the SOBP beams with high backgrund offset counts in both depth and lateral profiles, but they were quite noisy.
Fig. 4Depth (A) and lateral (B) profiles of prompt X-ray images with a 1.5-mm-diameter pinhole collimator of SOBP carbon-ion beams without a range shifter (RS: 0 cm), with a 2-cm range shifter (RS: 2 cm), and with a 4-cm range shifter (RS: 4 cm).
Energy spectra with a 1.5-mm pinhole collimator of prompt X-ray images of SOBP carbon-ion beams in the entire FOV are shown in Fig. 5. The energy spectra show a broad distribution with large peaks at 59 keV from tungsten-characteristic X-rays. The counts in the images did not change so much as the range shifter’s thickness increased.
Fig. 5Energy spectra of prompt X-ray images of SOBP carbon-ion beams in the entire FOV with a 1.5-mm diameter pinhole collimator.
3.2 Prompt X-ray images of SOBP beams with 4-mm-diameter collimators
Fig. 6(A), (B), and (C) show prompt X-ray images of a water phantom with a 4-mm-diameter pinhole collimator during irradiation by SOBP carbon-ion beams without a range shifter, with a 2-cm range shifter, and with a 4-cm range shifter, respectively. Here, the beam shapes can be observed in the images.
Fig. 6Prompt X-ray images with a 4-mm-diameter pinhole collimator for carbon-ion SOBP beams without a range shifter (A), with a 2-cm range shifter (B), and with a 4-cm range shifter (C).
Fig. 7(A) and (B) show depth and lateral profiles, respectively, of prompt X-ray images of a water phantom with a 4-mm-diameter pinhole collimator during irradiation of carbon-ion SOBP beams. We could observe the shapes of the SOBP beams with background offset counts in the depth and lateral profiles with some statistical noise.
Fig. 7Depth (A) and lateral (B) profiles of prompt X-ray images with a 4-mm-diameter pinhole collimator of carbon-ion SOBP beams without a range shifter (RS: 0 cm), with a 2-cm range shifter (RS: 2 cm), and with a 4-cm range shifter (RS: 4 cm).
Fig. 8(A), (B), and (C) show smoothed and background-subtracted prompt X-ray images from the images shown in Fig. 6(A), (B), and (C), respectively. Gaussian smoothing at 5 pixels was applied to the images in Fig. 6 to reduce the statistical noise. The background counts were subtracted by calculating the counts in the peripheral areas of the images. Here, clear shapes of the SOBP beams can be observed in the images, although the shapes were nearly circular due to the limited spatial resolution of the X-ray camera (∼50 mm FWHM at 40 cm).
Fig. 8Smoothed and background-subtracted prompt X-ray images with a 4-mm-diameter pinhole collimator for carbon-ion SOBP beams without a range shifter (A), with a 2-cm range shifter (B), and with a 4-cm range shifter (C).
Fig. 9(A) and (B) show the depth and lateral profiles of these smoothed and background-subtracted images. We could observe the shapes of the SOBP beams in both types of profiles.
Fig. 9Depth (A) and lateral (B) profiles of smoothed and background-subtracted prompt X-ray images with a 4-mm-diameter pinhole collimator for carbon-ion SOBP beams without a range shifter (RS: 0 cm), with a 2-cm range shifter (RS: 2 cm), and with a 4-cm range shifter (RS: 4 cm).
The ranges estimated at 10 % of the peak counts in Fig. 9(A) are summarized in Table 1 with calculated ranges. The calculated ranges were obtained by the dose delivery system of the carbon therapy system. The errors between the measured and calculated values were within 15 mm and 25 mm for the 4-mm and 1.5-mm pinhole collimators, respectively. The lateral widths estimated in FWHM are summarized in Table 2. The widths had almost the same values and were independent of the range shifter’s thickness for the 4-mm pinhole collimator.
Table 1Ranges estimated at 10 % of the peak counts measured with 4-mm and 1.5 mm diameter pinhole collimators with calculated ranges for carbon-ion SOBP beams.
Energy spectra of prompt X-ray images for carbon-ion SOBP beams with a 4-mm pinhole collimator are shown for the entire FOV in Fig. 10. The energy spectra showed a broad distribution with large peaks at 59 keV from tungsten-characteristic X-rays. The counts in the images decreased as the thickness of the range shifter increased.
Fig. 10Energy spectra of prompt X-ray images of carbon-ion SOBP beams with a 4-mm diameter pinhole collimator in the entire FOV.
Video clips of time sequential images of SOBP beams measured with a 4-mm pinhole collimator with a 100-ms interval are shown for range shifters of 0 cm, 2 cm, and 4 cm in Supplemental Materials 1, 2, and 3, respectively. These images were smoothed with a 2-point Gaussian filter. We could observe clear spills that appeared roughly every 2 s in each video.
The count rate curves measured with a pinhole of 4 mm in diameter from the 100-ms-interval images are shown in Fig. 11(A). The spills are repeated at a frequency of 0.5 Hz, and the width of the spills was ∼ 0.6 s. The time count rate curves of the 1-ms-interval images are shown in Fig. 11(B). The ripples were observed and repeated with a frequency of 0.3 kHz.
Fig. 11Count rate curves measured with a 4-mm-diameter pinhole collimator from 100-ms-interval images (A) and 1-ms-interval images (B).
We successfully acquired prompt X-ray images during irradiation of a water phantom with carbon-ion SOBP beams using a 4-mm-diameter pinhole collimator. Since the SOBP beams used for the imaging were at clinical dose levels, the images measured in this experiment provide evidence that the developed YAP(Ce) X-ray camera can be used for clinical imaging in carbon-ion therapy.
The background counts were high with carbon-ion SOBP beams due to the scattered prompt gamma photons from the MLC, so the imaging of SOBP beams was much more difficult than that of pencil beams using the developed pinhole X-ray camera. Recent particle therapy systems employ a spot-scanning method that does not require an MLC, which is expected to reduce the background fraction and provide much higher contrast images when these systems use SOBP or patient measurements. Thus the images measured in these experiments would represent the worst-case condition for SOBP or clinical prompt X-ray imaging.
One possible method to reduce the background fractions in the prompt X-ray images is to use the energy information. Since the peak areas of the tungsten-characteristic X-rays are mainly background counts, we may be able to reduce the background counts by forming the images using only energy that is lower than ∼ 50 keV. However reducing the energy window also decreases the counts from higher energy prompt X-rays, so some more experiments will be required to confirm that it is really effective.
In the estimated ranges for 4 mm diameter pinhole collimator listed in Table 1, the difference from the calculated range was largest with 4 cm range shifter. One possible explanation was that the fraction of secondary fragmentation particles by carbon ions was increased by the thick range shifter and the range was increased by these secondary fragmentation particles that have longer ranges. However more investigation will be required to confirm it is really the reason.
The energy spectra with a 1.5-mm-diameter pinhole collimator did not change so much with the thickness of the range shifter in Fig. 5. This was because the fraction of the prompt X-rays from the phantom was small. With the 4-mm-diameter collimator, the difference in the energy spectra was larger because the prompt X-ray fraction increased.
Another possible method to reduce the background fractions in the prompt X-ray images is to use a pinhole collimator with a larger diameter. A pinhole diameter larger than 4 mm may be effective in reducing the background fraction as well as increasing the sensitivity, though with a sacrifice in spatial resolution. The development of a new pinhole X-ray camera with a larger magnification ratio may compensate for the reduced spatial resolution of the X-ray camera with a larger-diameter pinhole collimator.
Time sequential images with 100-ms intervals as well as 1-ms intervals could be obtained with a 4-mm-diameter pinhole collimator. From the 100-ms-interval time sequential images, spills could be observed clearly. Moreover, from the 1-ms-interval time sequential images, ripples could be clearly observed, although the images showed low counts. These images might be useful for checking the temporal accuracy of the beams for future gated particle therapy or FLASH radiotherapy [
Reduced cognitive deficits after FLASH irradiation of whole mouse brain are associated with less hippocampal dendritic spine loss and neuroinflammation.
Darafsheh A, Hao Y, Zwart X, Wagner M, Catanzano D, Williamson JF, et al., Feasibility of proton FLASH irradiation using a synchrocyclotron for preclinical studies, Med Phys. 42020; 7, 4348-435.
The distribution of the prompt X-ray images was different from the dose distributions of the carbon-ion beams for SOBP beams. This is because the production of prompt X-rays is small for the low-energy secondary electrons produced by the particle ions as well as for the larger attenuation of the lower-energy prompt X-rays at the end of the beams. However, as we previously showed, the distributions could be corrected to the dose distribution by the use of a neural network employing deep learning approaches [
The developed pinhole YAP(Ce) X-ray camera combined with a 4-mm-diameter pinhole collimator enabled the imaging of SOBP beams at clinical dose levels with an MLC, in which the counts were low and background levels were high. The SOBP images obtained in this experiment serve as evidence that our prompt X-ray imaging system can be used in clinical conditions.
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgements
This work was supported by JST ERATO Grant Number JPMJER2102, Japan. This work was also partly supported by JSPS KAKENHI Grant Number 19H00672 and 22H03019.
Patient study on in-vivo verification of beam delivery and range using positron emission tomography and computed tomography imaging after proton therapy.
M. Yamaguchi, Torikai K, Kawachi N, Shimada H, Satoh T, Y. Nagao Y, et al., Beam range estimation by measuring bremsstrahlung, Phys Med Biol. 2012; 57,2843-56.
Development of a low-energy x-ray camera for the imaging of secondary electron bremsstrahlung x-ray emitted during proton irradiation for range estimation.
Yamaguchi M, Nagao Y, Ando K, Yamamoto S, Sakai M, R. Parajuli RK, et al., Imaging of monochromatic beams by measuring secondary electron bremsstrahlung for carbon-ion therapy using a pinhole x-ray camera, Phys Med Biol. 2018; 63(4),045016.
Development of a YAP(Ce) camera for the imaging of secondary electron bremsstrahlung x-ray emitted during carbon-ion irradiation toward the use of clinical conditions.
Sensitivity improvement of YAP(Ce) cameras for imaging of secondary electron bremsstrahlung X-rays emitted during carbon-ion irradiation: Problem and solution.
Yamamoto S, Watabe H, Kanai Y, Watabe T, Kato K, J. Hatazawa J. Development of an ultrahigh resolution Si-PM based PET system for small animals. Phys Med Biol. 2013; 58,7875-88.
Reduced cognitive deficits after FLASH irradiation of whole mouse brain are associated with less hippocampal dendritic spine loss and neuroinflammation.
Darafsheh A, Hao Y, Zwart X, Wagner M, Catanzano D, Williamson JF, et al., Feasibility of proton FLASH irradiation using a synchrocyclotron for preclinical studies, Med Phys. 42020; 7, 4348-435.
00069-8&id=gr1.jpg){kind=link}
00069-8&id=gr2.jpg){kind=link}
00069-8&id=gr3.jpg){kind=link}
00069-8&id=gr4.jpg){kind=link}
00069-8&id=gr5.jpg){kind=link}
00069-8&id=gr6.jpg){kind=link}
00069-8&id=gr7.jpg){kind=link}
00069-8&id=gr8.jpg){kind=link}
00069-8&id=gr9.jpg){kind=link}
00069-8&id=gr10.jpg){kind=link}
00069-8&id=gr11.jpg){kind=link}