Keisuke Usui

Dose calculation with a cone beam CT image in image-guided radiation therapy

A kilo-voltage cone-beam CT (CBCT) attached to a linear accelerator can verify a target position in each radiation therapy. If CBCT images can be used in dose calculation, we can verify an actual dose distribution on every treatment day. However, the CBCT images are degraded by several factors, and so we cannot use the CBCT images directly in place of conventional multi-slice CT (MSCT) images that are used in the initial dose planning. In this paper, we proposed a new method for using CBCT and MSCT images in the calculation of a dose distribution. Our proposed method segments the CBCT and MSCT images into regions of three major organs (lungs, bones and soft tissues) by use of histogram analysis. We also calculated a value such as the median of the MSCT numbers in each region of the MSCT images, and we set three representative values to the corresponding regions of the CBCT images. In the calculation of a dose distribution, we used these modified CBCT images. The validity of our method was confirmed with experiments in which we used images of a heterogeneous phantom and patients’ lungs in comparison with conventional methods. The results showed that the dose distribution determined by our method was similar to that of the initial dose plan, and our method was superior to the conventional methods in terms of pass rates of a distance-to-agreement analysis and γ analysis. The results of a dose-volume-histogram analysis also showed the accuracy of our proposed method.

Effect of region extraction and assigned mass-density values on the accuracy of dose calculation with magnetic resonance-based volumetric arc therapy planning

This study aimed to verify the validity of generating treatment plans for volumetric arc therapy (VMAT) for prostate cancer using magnetic resonance (MR) imaging with a dose calculation algorithm in Acuros XB (Eclipse version 13.6; Varian Medical Systems, Palo Alto, CA, USA) based on deterministically solving the linear Boltzmann transport equations. Four different classes were applied to prostate MR images: MRW (all water equivalent); MRW+B (water and bone); MRS+B (soft tissue and bone); and MRS+B+G (soft tissue, bone, and rectal gas). Each of these regions was assigned a mass density for calculating doses. The assigned mass-density values were then altered in three ways. Using initial planning and optimization parameters, MR-based VMAT plans were generated and compared with corresponding forward-calculated computed tomography-based plans for doses to the target volumes and organs at risk using dose-volume histograms and γ analyses. In the MRW plans, the mean doses for TVs were overestimated by approximately 1.3%. The MRW+B plans revealed reduced differences within 0.5%. Further segmentation (MRS+B) did not result in substantial improvement. Dose deviations affected by the changes in the mass densities assigned to soft tissue were as small as approximately 1.0%, whereas larger deviations were revealed in bone and rectal gas, especially those with > 5% error. Assignment of accurate mass-density values acquired from MR images is needed for MR-based radiation treatment planning. Multiple MR sequences should be acquired for segmentation and mass-density conversion purposes. Segmented MR-based VMAT planning is feasible with a density assignment method using Acuros XB.

[Impact of the Infrared Monitor Signal Pattern on Accuracy of Target Imaging in 4-dimensional Cone-beam Computed Tomography].

To realize the high precision radiotherapy, localized radiation field of the moving target is very important, and visualization of a temporal location of the target can help to improve the accuracy of the target localization. However, conditions of the breathing and the patients own motion differ from the situation of the treatment planning. Therefore, positions of the tumor are affected by these changes. In this study, we implemented a method to reconstruct target motions obtained with the 4D CBCT using the sorted projection data according to the phase and displacement of the extracorporeal infrared monitor signal, and evaluated the proposed method with a moving phantom. In this method, motion cycles and positions of the marker were sorted to reconstruct the image, and evaluated the image quality affected by changes in the cycle, phase, and positions of the marker. As a result, we realized the visualization of the moving target using the sorted projection data according to the infrared monitor signal. This method was based on the projection binning, in which the signal of the infrared monitor was surrogate of the tumor motion. Thus, further major efforts are needed to ensure the accuracy of the infrared monitor signal.

Visualization of a target positions using the 4 dimensional cone-beam CT image reconstruction with the extracorporeal infrared monitor

To determine the accurate radiation dose for a specific small volume of tissue, confirmation of the temporal location of a target position is very important. The temporal location of the target can help improve the accuracy of target localization. A kilovoltage cone-beam computed tomography (CBCT) system mounted on a linear accelerator can verify the target on each treatment day, but this system requires a long data acquisition time; therefore, reconstructed images are affected by motion. In this study, we implemented a method to reconstruct target motions obtained by a four-dimensional CBCT using sorted projection data according to the phase and displacement of the extracorporeal infrared monitor signal. We evaluated the effect of reconstruction methods on image quality using a moving phantom and respiratory signals of an actual patient. We modified the relationship between the motion cycle and positions of the marker and projection data and evaluated the effect on image quality. Moreover, the effect of target motions in patients on the quality of reconstructed images was investigated. The results of the experiments showed that phase binning and displacement binning reduced the blurring of the reconstructed image. The quality of the reconstructed images was significantly affected by the amplitude as well as the inhalation and exhalation slopes of the marker signal. The missing number of projections in the displacement binning method was caused by non-linearity of breathing pattern. This method was based on projection binning, in which the signal of the infrared monitor served as the surrogate of tumor motion. Therefore, further major efforts are needed to ensure the accuracy of the infrared monitor signal.

SU-E-T-465: Dose Calculation Method for Dynamic Tumor Tracking Using a Gimbal-Mounted Linac

PURPOSE Dynamic tumor tracking using the gimbal-mounted linac (Vero4DRT, Mitsubishi Heavy Industries, Ltd., Japan) has been available when respiratory motion is significant. The irradiation accuracy of the dynamic tumor tracking has been reported to be excellent. In addition to the irradiation accuracy, a fast and accurate dose calculation algorithm is needed to validate the dose distribution in the presence of respiratory motion because the multiple phases of it have to be considered. A modification of dose calculation algorithm is necessary for the gimbal-mounted linac due to the degrees of freedom of gimbal swing. The dose calculation algorithm for the gimbal motion was implemented using the linear transformation between coordinate systems. METHODS The linear transformation matrices between the coordinate systems with and without gimbal swings were constructed using the combination of translation and rotation matrices. The coordinate system where the radiation source is at the origin and the beam axis along the z axis was adopted. The transformation can be divided into the translation from the radiation source to the gimbal rotation center, the two rotations around the center relating to the gimbal swings, and the translation from the gimbal center to the radiation source. After operating the transformation matrix to the phantom or patient image, the dose calculation can be performed as the no gimbal swing. The algorithm was implemented in the treatment planning system, PlanUNC (University of North Carolina, NC). The convolution/superposition algorithm was used. The dose calculations with and without gimbal swings were performed for the 3 × 3 cm2 field with the grid size of 5 mm. RESULTS The calculation time was about 3 minutes per beam. No significant additional time due to the gimbal swing was observed. CONCLUSIONS The dose calculation algorithm for the finite gimbal swing was implemented. The calculation time was moderate.

CBCT image reconstruction of a moving target with an on-board imaging system for radiation therapy

To accomplish accurate irradiation to a target, image information about the target is very useful in radiotherapy. A kilo-voltage cone-beam computed tomography (kV-CBCT) system mounted on a linear accelerator can verify the accuracy of the set-up position and the size and location of the target on each treatment day. However, the gantry of the CBCT system is heavy and so requires a long data acquisition time. As a result, image distortion occurs in the case of a moving target. In this study, we proposed a method to reconstruct a target less affected by any motion of the target, and evaluated the proposed method with a moving phantom. In our method, CBCT images of the moving phantom were acquired, and the target position in two directions was detected with a template matching method using these projection images. And we selected data acquisition angles of the projection images in which the target was located in predefined regions in the projection images, and made a sinogram with the projection image of selected angles. A less blurred image was reconstructed with a filtered-backprojection method from the sinogram. The results of experiments showed that the blurring caused by the movement of the phantom was reduced with our proposed method.

Evaluation on Pixel Value Conversion Methods for Dose Calculation with Kilo-Voltage Cone Beam CT Images

The purpose of our study is to evaluate the accuracy of pixel-value conversion methods for dose calculation with kilo-voltage cone-beam computed tomography (CBCT) images in radiotherapy. We investigated the relationship between CT numbers reconstructed with a CBCT and a conventional multi- slice CT (MSCT). To convert CBCT numbers to MSCT numbers, we used three conversion methods. One was the method using an electron-density CT-phantom (EP method). The second was the method using mean values in arbitrary defined several specific areas in CBCT and MSCT images (SA method). The last was the method using segmented regions (SR method), in which we extracted lung, bone and soft-tissue regions in CBCT and MSCT images, and replaced the pixel value of the CBCT image with the median (or mean, mode) of the MSCT numbers in each corresponding region. And after the conversion of CBCT numbers to MSCT numbers, we calculated the dose distribution using this image. The accuracy of dose distribution was verified with a heterogeneous phantom and clinical lung images. We compared the results with initial plans using a pass rate. The differences in distance-toagreement (DTA) 2mm pass rates between the dose distribution calculated with a converted CBCT image and that of the initial plan of the phantom were 98.5%, 100%, 100% and 20.4% for the EP, SA, SR method and the case without any conversion, respectively. And in the results of the γ analysis (2mm, 2%), the pass rates were 100%, 100%, 100% and 30.5% for the EP, SA, SR method and the case without any conversion, respectively. In the SR method (median) the accuracy was much improved. As for the clinical lung images, the same tendency was observed as that of the phantom. The SR method could improve the dose calculation accuracy using a converted CBCT image.

SU‐E‐J‐100: Feasibility of Dose Calculation Using Combined Information of Cone‐Beam and Multi‐Slice CT Images

PURPOSE The purpose of this study is to evaluate the feasibility of kilovoltage cone-beam CT (CBCT) images that are obtained with the Varian On-Board Imager in dose calculation at each radiation therapy. METHODS CBCT images are commonly degraded by scattered radiations originating in the patients body, and so the CT numbers of the CBCT images depend on data acquisition conditions and the patient size. However, the anatomical shape of each organ is not likely affected by scattered radiations, and so we used only the shape of major organs such as lungs and bones in the CBCT images, and replaced these CT numbers with those of the multi-slice CT (MSCT) images that were used for dose calculation in a treatment planning. As regards this alternative CT number we adopted the median of MSCT numbers in a segmented region of a major organ each corresponding to that in the CBCT images. We evaluated the validity of our segmented region (SR) method with images of eight patients with lung diseases. The number of irradiation beams was four. In this evaluation we used the distance-to-agreement (DTA) and y analysis, and the dose-volume-histogram (DVH) analysis. RESULTS The pass rates of the DTA analysis (2mm) and γ analysis (2mm, 2%) between the dose distributions calculated with our method were 90.4±6.0% and 99.1±1.1%, respectively. The results of the DVH analysis showed that the differences in doses (average, maximum and minimum) for a target volume were 1.3±0.5%, 0.9±0.8% and 3.4±3.0%, respectively. These results showed that our method was acceptable in the calculation of a dose distribution. CONCLUSION We evaluated the dose calculation method with a combination of CBCT and MSCT images. This method could yield an accurate dose distribution and achieved an easier verification of radiation therapy on each treatment day.