- Email: [email protected]
Respiratory Gating to Minimize the Organ Motion Effect
S558
I. J. Radiation Oncology
● Biology ● Physics
Volume 63, Number 2, Supplement, 2005
Figure 1. Spinal cord and Level I Lymph Node dose reconstruction for the initial 25 fractions (50 Gy). Note that the spinal cord (A) is correctly aligned between the daily CT (yellow) and the planning CT (gray), but the Level I Nodes are not aligned (B).
2545
Combining Cardiac/Respiratory Gating to Minimize the Organ Motion Effect C. Willett, Z. Wang, L. Marks, T. Raidy, K. Kelly, M. Oldham, S. Das, S. Zhou, M. Kasibhalta, F. Yin Radiation Oncology, Duke University Medical Center, Durham, NC Purpose/Objective: Respiratory gating techniques have been demonstrated to facilitate margin reductions by effectively reducing organ motion due to respiration. We investigate the feasibility to reduce organ motion further by combining simultaneous cardiac and respiratory gating techniques. Materials/Methods: Four human subjects were serially scanned in the supine position with a GE 1.5T MRI scanner. T1-weighted sagittal and coronal localization scans were first performed to define the regions to be studied. Axial images through the heart, liver and pancreas were continuously generated with fast cine MRI scans at three different gating settings: a). without either respiratory or cardiac gating, b). with respiratory gating but without cardiac gating, and c). with both respiratory and cardiac gating. The motions of heart, liver and pancreas were calculated at the organ edges with the maximum motions along the direction that is perpendicular to the edge. Results: For the scans without either respiratory or cardiac gating, the detected maximum organ motion on the axial images averaged among four subjects was 1.4 cm for heart, 1.1 cm for liver, and 1.0 cm for pancreas. For the scans with respiratory gating but without cardiac gating, the maximum organ motion on the axial images averaged among four subjects was 0.7 cm for heart, 0.5 cm for liver, and 0.4 cm for pancreas. For the scans with both respiratory and cardiac gating, the motion was ⬍0.2 cm for heart, liver and pancreas. The movie files were also created and showed significant differences in organ motion for three different gating settings. Conclusions: Significant organ motion is still present in heart, liver and pancreas even when respiratory gating is applied. In some clinical settings, the application of both cardiac and respiratory gating may be therapeutically advantageous. Additional study is warranted to better understand this issue.
Proceedings of the 47th Annual ASTRO Meeting
Figure 1: (a-c): The difference image (subtraction between two images acquired at different times) of an image through the heart: a) Without either respiratory or cardiac gating, b) With respiratory gating but without cardiac gating. Motion reduced but still present, c) With both respiratory and cardiac gating. Very minimum motion present. The residual signals came from blood, and (d): The anatomic image without subtraction.
2546
Integral Radiation Dose to Normal Structures with Conformal External Beam Radiation
H. Aoyama,1,4 D.C. Westerly,2 T.R. Mackie,2,3 S.M. Bentzen,1 G.H. Olivera,3,2 R.R. Patel,1 H. Jaradat,1 W.A. Tome,1 M.A. Ritter,1 M.P. Mehta1 1 Human Oncology, University of Wisconsin, Madison, WI, 2Medical Physics, University of Wisconsin, Madison, WI, 3 TomoTherapy Inc, Madison, WI, 4Radiology, Hokkaido University Graduate School of Medicine, Sapporo, Japan Purpose/Objective: Integral dose (ID) is the volume integral of the dose deposited in a patient and is equal to the mean dose times the volume irradiated to any dose. It is often claimed that the large number of beamlets and monitor units used in IMRT leads to an increase in ID, and that higher energy photon beams substantially reduce the normal tissue ID (NTID), but these claims have rarely been systematically studied. This dosimetric study was designed to evaluate 25 different treatment plans, specifically comparing the NTID from intensity-modulated radiotherapy (IMRT), as well as 3DCRT for prostate cancer. The risk of secondary malignancies was also estimated based on the dose-volume histograms (DVH). Materials/Methods: Twenty-five radiation treatment plans, including IMRT using a conventional linac with both 6 (6MVIMRT) and 20 MV (20MV-IMRT) as well as 3DCRT using 6 (6MV-3DCRT) and 20 MV (20MV-3DCRT), and IMRT using Tomotherapy, were created for five consecutive patients with localized prostate cancer. A total of 70Gy was prescribed to encompass 95 % of the PTV. The ID to normal tissue was calculated from the DVH. The risk of secondary malignancies was also estimated using a biological model. The epidemiological study by Brenner et al. (Cancer 2000; 88: 398 – 406) found an excess risk of sarcomas and cancer of the lung, bladder, and rectum in 51,000 men receiving radiotherapy compared with 71,000 men undergoing radical prostatectomy. Applying these figures and the radiation-dose / risk estimations from Hall and Wuu (IJROBP 2003; 56: 83– 88) to the calculated dose distributions from our study, provided an estimate of the relative number of excess cancer cases for the 5 delivery methods. Results: The percent difference of the ID for the PTV of each plan was ⫾ 0.2%, compared with 6MV-3DCRT; therefore all plans were clinically equivalent in terms of PTV coverage. The mean NTID was 122.8 liter-Gy for 6MV-3DCRT, 116.6 for 6MV-IMRT, 113.3 for 20MV-3DCRT, 109.0 for 20MV-IMRT, and 117.8 for Tomotherapy. Compared to 6MV-3DCRT, the percent difference of 6MV-IMRT and Tomotherapy was ⫺5.2% and ⫺4.1% respectively. The NTIDs of 20MV-3DCRT and 20MV-IMRT were smaller than 6MV-3DCRT and 6MV-IMRT by 8% and 6.7%, respectively. Regarding the IDs of rectum, penile bulb, and bladder, the use of 20MV did not reduce the ID to these structures. Compared to 6MV-3DCRT, the percent difference in the IDs of 20MV-3DCRT / 20MV-IMRT were 0%/⫺1.8% for rectum, 4.7%/⫺0.3% for penile bulb, and ⫺2.9%/⫺0.7% for bladder. The use of 6MV-IMRT reduced the ID only modestly, by ⫺6.8% for the rectum, ⫺2.8% for the penile bulb, and ⫺7.8% for the bladder whereas Tomotherapy resulted in somewhat greater reduction of ⫺12.6% for the rectum, ⫺18% for the penile bulb, and ⫺6.2% for the bladder. The five methods gave estimates of secondary malignancies that varied within 0.1 % for bladder, rectum, and sarcoma. This variation appears to be an order of magnitude lower than previous estimates, which did not incorporate the vigor of definitive DVH based NTID computation. Secondary malignancy estimates show minimal variations between techniques. Dosimetric measurement for lung is in progress and will be presented at the meeting. Conclusions: The differences between 3DCRT and IMRT in terms of NTID are very small and may even be negligible if detailed leakage and neutron dose determination are accounted for. The advantage of Tomotherapy over conventional IMRT and 3DCRT for localized prostate cancer was demonstrated in regard to dose sparing of rectum and penile bulb without increasing NTID and the risk of secondary malignancy.
2547
A 5-Dimensional Model of Human Breathing Motion 1
D.A. Low, W. Lu,1 P.J. Parikh,1 J.F. Dempsey,2 S. Mutic,1 J.D. Bradley1 Department of Radiation Oncology, Washington University, St. Louis, MO, 2Radiation Oncology, University of Florida, Gainesville, FL 1
Purpose/Objective: A mathematical model of human breathing motion has been developed for radiation therapy treatment planning. Materials/Methods: The model relates the positions of all regions in lungs (normal and tumor tissues) as functions of tidal volume and airflow. Characterization of position as functions of tidal volume and airflow model position as a function of breathing depth and allow for a quantitative hysteresis description, respectively. The model therefore uses 5 intrinsic degrees-of-freedom (three coordinates at a reference breathing phase, tidal volume and airflow). The time-dependence of position (motion) comes from the time dependence of the tidal volume and airflow. A linear approximation of this model is being evaluated with 4D CT lung images, where the position of regions relative to the reference breathing phase are composed of two vectors, one whose length is proportional to tidal volume and one proportional to airflow. The orientations and relative magnitudes of these vectors vary throughout the lungs and tumors to provide the complex, deformable motion known to exist within human breathing. The 4D CT process used a 16-slice CT scanner (Philips Brilliance) operating in cine mode with concurrent spirometry, yielding 25 16-slice CT scans for each couch position with sufficient couch positions acquired to cover the lungs. Regions are automatically selected and tracked in a 5 mm ⫻ 5 mm grid centered in each couch position. Each patient has between 1000 and 2000 regions tracked. The region’s locations, tidal volumes, and airflows are fit to the motion model and the parameters of the model smoothed using a 5 mm standard deviation Gaussian filter. For presentation of the data, the predicted positions are presented as vectors from tidal inhalation to exhalation.
S559
I. J. Radiation Oncology
● Biology ● Physics
Volume 63, Number 2, Supplement, 2005
Figure 1. Spinal cord and Level I Lymph Node dose reconstruction for the initial 25 fractions (50 Gy). Note that the spinal cord (A) is correctly aligned between the daily CT (yellow) and the planning CT (gray), but the Level I Nodes are not aligned (B).
2545
Combining Cardiac/Respiratory Gating to Minimize the Organ Motion Effect C. Willett, Z. Wang, L. Marks, T. Raidy, K. Kelly, M. Oldham, S. Das, S. Zhou, M. Kasibhalta, F. Yin Radiation Oncology, Duke University Medical Center, Durham, NC Purpose/Objective: Respiratory gating techniques have been demonstrated to facilitate margin reductions by effectively reducing organ motion due to respiration. We investigate the feasibility to reduce organ motion further by combining simultaneous cardiac and respiratory gating techniques. Materials/Methods: Four human subjects were serially scanned in the supine position with a GE 1.5T MRI scanner. T1-weighted sagittal and coronal localization scans were first performed to define the regions to be studied. Axial images through the heart, liver and pancreas were continuously generated with fast cine MRI scans at three different gating settings: a). without either respiratory or cardiac gating, b). with respiratory gating but without cardiac gating, and c). with both respiratory and cardiac gating. The motions of heart, liver and pancreas were calculated at the organ edges with the maximum motions along the direction that is perpendicular to the edge. Results: For the scans without either respiratory or cardiac gating, the detected maximum organ motion on the axial images averaged among four subjects was 1.4 cm for heart, 1.1 cm for liver, and 1.0 cm for pancreas. For the scans with respiratory gating but without cardiac gating, the maximum organ motion on the axial images averaged among four subjects was 0.7 cm for heart, 0.5 cm for liver, and 0.4 cm for pancreas. For the scans with both respiratory and cardiac gating, the motion was ⬍0.2 cm for heart, liver and pancreas. The movie files were also created and showed significant differences in organ motion for three different gating settings. Conclusions: Significant organ motion is still present in heart, liver and pancreas even when respiratory gating is applied. In some clinical settings, the application of both cardiac and respiratory gating may be therapeutically advantageous. Additional study is warranted to better understand this issue.
Proceedings of the 47th Annual ASTRO Meeting
Figure 1: (a-c): The difference image (subtraction between two images acquired at different times) of an image through the heart: a) Without either respiratory or cardiac gating, b) With respiratory gating but without cardiac gating. Motion reduced but still present, c) With both respiratory and cardiac gating. Very minimum motion present. The residual signals came from blood, and (d): The anatomic image without subtraction.
2546
Integral Radiation Dose to Normal Structures with Conformal External Beam Radiation
H. Aoyama,1,4 D.C. Westerly,2 T.R. Mackie,2,3 S.M. Bentzen,1 G.H. Olivera,3,2 R.R. Patel,1 H. Jaradat,1 W.A. Tome,1 M.A. Ritter,1 M.P. Mehta1 1 Human Oncology, University of Wisconsin, Madison, WI, 2Medical Physics, University of Wisconsin, Madison, WI, 3 TomoTherapy Inc, Madison, WI, 4Radiology, Hokkaido University Graduate School of Medicine, Sapporo, Japan Purpose/Objective: Integral dose (ID) is the volume integral of the dose deposited in a patient and is equal to the mean dose times the volume irradiated to any dose. It is often claimed that the large number of beamlets and monitor units used in IMRT leads to an increase in ID, and that higher energy photon beams substantially reduce the normal tissue ID (NTID), but these claims have rarely been systematically studied. This dosimetric study was designed to evaluate 25 different treatment plans, specifically comparing the NTID from intensity-modulated radiotherapy (IMRT), as well as 3DCRT for prostate cancer. The risk of secondary malignancies was also estimated based on the dose-volume histograms (DVH). Materials/Methods: Twenty-five radiation treatment plans, including IMRT using a conventional linac with both 6 (6MVIMRT) and 20 MV (20MV-IMRT) as well as 3DCRT using 6 (6MV-3DCRT) and 20 MV (20MV-3DCRT), and IMRT using Tomotherapy, were created for five consecutive patients with localized prostate cancer. A total of 70Gy was prescribed to encompass 95 % of the PTV. The ID to normal tissue was calculated from the DVH. The risk of secondary malignancies was also estimated using a biological model. The epidemiological study by Brenner et al. (Cancer 2000; 88: 398 – 406) found an excess risk of sarcomas and cancer of the lung, bladder, and rectum in 51,000 men receiving radiotherapy compared with 71,000 men undergoing radical prostatectomy. Applying these figures and the radiation-dose / risk estimations from Hall and Wuu (IJROBP 2003; 56: 83– 88) to the calculated dose distributions from our study, provided an estimate of the relative number of excess cancer cases for the 5 delivery methods. Results: The percent difference of the ID for the PTV of each plan was ⫾ 0.2%, compared with 6MV-3DCRT; therefore all plans were clinically equivalent in terms of PTV coverage. The mean NTID was 122.8 liter-Gy for 6MV-3DCRT, 116.6 for 6MV-IMRT, 113.3 for 20MV-3DCRT, 109.0 for 20MV-IMRT, and 117.8 for Tomotherapy. Compared to 6MV-3DCRT, the percent difference of 6MV-IMRT and Tomotherapy was ⫺5.2% and ⫺4.1% respectively. The NTIDs of 20MV-3DCRT and 20MV-IMRT were smaller than 6MV-3DCRT and 6MV-IMRT by 8% and 6.7%, respectively. Regarding the IDs of rectum, penile bulb, and bladder, the use of 20MV did not reduce the ID to these structures. Compared to 6MV-3DCRT, the percent difference in the IDs of 20MV-3DCRT / 20MV-IMRT were 0%/⫺1.8% for rectum, 4.7%/⫺0.3% for penile bulb, and ⫺2.9%/⫺0.7% for bladder. The use of 6MV-IMRT reduced the ID only modestly, by ⫺6.8% for the rectum, ⫺2.8% for the penile bulb, and ⫺7.8% for the bladder whereas Tomotherapy resulted in somewhat greater reduction of ⫺12.6% for the rectum, ⫺18% for the penile bulb, and ⫺6.2% for the bladder. The five methods gave estimates of secondary malignancies that varied within 0.1 % for bladder, rectum, and sarcoma. This variation appears to be an order of magnitude lower than previous estimates, which did not incorporate the vigor of definitive DVH based NTID computation. Secondary malignancy estimates show minimal variations between techniques. Dosimetric measurement for lung is in progress and will be presented at the meeting. Conclusions: The differences between 3DCRT and IMRT in terms of NTID are very small and may even be negligible if detailed leakage and neutron dose determination are accounted for. The advantage of Tomotherapy over conventional IMRT and 3DCRT for localized prostate cancer was demonstrated in regard to dose sparing of rectum and penile bulb without increasing NTID and the risk of secondary malignancy.
2547
A 5-Dimensional Model of Human Breathing Motion 1
D.A. Low, W. Lu,1 P.J. Parikh,1 J.F. Dempsey,2 S. Mutic,1 J.D. Bradley1 Department of Radiation Oncology, Washington University, St. Louis, MO, 2Radiation Oncology, University of Florida, Gainesville, FL 1
Purpose/Objective: A mathematical model of human breathing motion has been developed for radiation therapy treatment planning. Materials/Methods: The model relates the positions of all regions in lungs (normal and tumor tissues) as functions of tidal volume and airflow. Characterization of position as functions of tidal volume and airflow model position as a function of breathing depth and allow for a quantitative hysteresis description, respectively. The model therefore uses 5 intrinsic degrees-of-freedom (three coordinates at a reference breathing phase, tidal volume and airflow). The time-dependence of position (motion) comes from the time dependence of the tidal volume and airflow. A linear approximation of this model is being evaluated with 4D CT lung images, where the position of regions relative to the reference breathing phase are composed of two vectors, one whose length is proportional to tidal volume and one proportional to airflow. The orientations and relative magnitudes of these vectors vary throughout the lungs and tumors to provide the complex, deformable motion known to exist within human breathing. The 4D CT process used a 16-slice CT scanner (Philips Brilliance) operating in cine mode with concurrent spirometry, yielding 25 16-slice CT scans for each couch position with sufficient couch positions acquired to cover the lungs. Regions are automatically selected and tracked in a 5 mm ⫻ 5 mm grid centered in each couch position. Each patient has between 1000 and 2000 regions tracked. The region’s locations, tidal volumes, and airflows are fit to the motion model and the parameters of the model smoothed using a 5 mm standard deviation Gaussian filter. For presentation of the data, the predicted positions are presented as vectors from tidal inhalation to exhalation.
S559