S740
International Journal of Radiation Oncology Biology Physics
Results: In three of the fractions (two arcs each) the positioning of the patient was good. At least one of the three fiducials was visible during almost entire treatment (95% and 85% of the images for each arc, respectively). In those images, the marker detection algorithm was accurate and agrees well with manual detection (andlt 0.2 mm). The mean displacement for all fiducials and fractions is 0.85 mm. Detailed information for each fiducial/fraction/arc is reported in the attached table. In the fourth fraction (one arc), the movement of the prostate was larger. The mean displacement with respect to its planned position is 2.85 0.55 mm. Conclusion: An algorithm for automatic detection of fiducial markers in cine MV images has been developed. It was used to analyze the prostate movement during a VMAT treatment for a patient with implanted markers. In light of the random nature of intrafraction prostate motion, this work represents an important step toward real-time image-guided prostate radiation therapy. Future work includes making the algorithm more robust and analyzing more patients. Author Disclosure: J. Azcona: None. R. Li: None. E. Mok: None. S. Hancock: None. L. Xing: None.
a 95% Confidence Interval (CI) of 7.0 mm, which is within the bounds of image guided shifts reported in the previous study. The mean SD of all the alignments for the same patient had a maximum of 2.0, 3.1, 2.1 mm in LR, AP, SI directions respectively. The small SD suggests that US shifts were consistently reproducible for users with appropriate training. Conclusions: We believe that our data indicates that US is a viable image guidance method for the treatment of abdominal tumors that can be straightforwardly implemented at multiple institutions. Author Disclosure: B. Wang: None. V. Sarkar: None. C.J. Anker: None. S. Streitmatter: None. P. Rassiah-Szegedi: None. H. Zhao: None. J.Y. Huang: None. M. Szegedi: None. B.J. Salter: None.
3388
Purpose/Objective(s): During Head and Neck treatment, some patients complain of claustrophobia due to the fully closed thermoplastic head masks. Recently introduced open mask solve this by exposing the patients’ eyes, nose, and mouth. This study investigates the immobilization capability of open masks and proposes new CTV to PTV expansion margins for them. Materials/Methods: Open masks were deformed to custom-fit two volunteers using “C” size head pillows on a CT. After a number of days, the volunteers found the most comfortable position on the Linac treatment couch in the mask. A 3-D skin tracking system acquired a reference surface per volunteer. An ROI consisting of the eyes, nose, mouth, and part of cheek was selected per volunteer. The volunteer’s setup ROI was tracked and rigidly registered to the reference ROI to calculate the skin surface’s translation and rotation relative to the reference surface. The mask was removed, and the volunteer was asked to sit up. The setup was repeated five times to simulate daily setup uncertainties. In a second test, volunteers tried to remain still in the mask for 10 minutes to simulate the duration of a typical treatment. We recorded the translational and rotational real time deltas and the maximum deltas. Results: The Results Table shows, per volunteer, the average deltas over the 5 setups and the corresponding standard deviations. We also report the maximum delta from reference observed during the 10-minute monitoring interval. We observed that the real time deltas increase slightly with time spent on the table, which is to be expected. Using a commonly accepted guideline (2.5average + 0.7standard deviation), we calculate the implied CTV to PTV expansion. Conclusions: Open masks generally solve claustrophobia. Based on our preliminary study, the conventional CTV to PTV expansion (3, 3, 3)mm should be altered to (2, 4, 4) mm for open mask usage. The open masks appear to provide sufficient immobilization. More volunteer study is scheduled to be done in the next several months. Author Disclosure: X. Tang: None. G. Tracton: None. R. Rupolo: None. Z. Xu: None. S. Wang: None. M. Lawrence: None.
Ultrasound Image Guidance for Abdominal Tumor B. Wang, V. Sarkar, C.J. Anker, S. Streitmatter, P. Rassiah-Szegedi, H. Zhao, J.Y. Huang, M. Szegedi, and B.J. Salter; University of Utah, Salt Lake City, UT Purpose/Objective(s): A newly implemented program of using ultrasound (US) as an image guidance tool for abdominal tumors has been established at our institution based on experience previously acquired by our staff at another institution. This study evaluated our initial experience and compared our initial image guided shift data to the previously published literature. The data should be useful in ascertaining if experience in performing US image guidance for abdominal tumors can be replicated at a new institution. Materials/Methods: A total of 414 fractions delivered to 15 patients treated for pancreatic and hepatobiliary carcinoma were retrospectively analyzed, with each patient receiving between 25 and 28 fractions. Patients were first aligned to skin marks with treatment room lasers. US image guidance was then performed to align relevant reference arterial and venous vasculature to their corresponding contours from the CT simulation using the BAT system (1). The image guidance shifts by US were recorded as the residual difference between setup on skin marks and setup with US. Results: Consistent with previously published work, following appropriate training of radiation therapists, vascular “reference” structures were readily identified for the image guidance alignments. The maximum of residual differences for all 414 fractions was 6.7, 9.6, and 6.6 mm in left-right (LR), anterior-posterior (AP), and superior-inferior (SI) directions, respectively. These maximum shifts were smaller than the previously reported maximum shifts values (w20, 30, 30 mm), but are likely due our smaller patient sample size. Nevertheless, such nontrivial maximum shifts emphasize the importance of image guidance for treatment of these abdominal tumors. The 3D magnitude vector had a mean of 2.2 mm and Standard Deviation (SD) of 2.4 mm, with
Poster Viewing Abstract 3389; Table
3389 Investigation of the Immobilization Capability of Open Masks for Head and Neck Patient Setup and Treatment X. Tang, G. Tracton, R. Rupolo, Z. Xu, S. Wang, and M. Lawrence; University of North Carolina, Chapel Hill, NC
Study results (real time deltas)
Volunteer
Metric
Vertical (mm)
Lateral (mm)
Longitudinal (mm)
Yaw (degrees)
Roll (degrees)
Pitch (degrees)
1 1 1 1 2 2 2 2
Average delta St. dev. of delta Max delta in 10 minutes Expansion Average delta St. dev. of delta Max delta in 10 minutes Expansion
-0.9 0.86 0.4 1.7 0.1 2.12 0.4 1.5
-1.7 1.19 0.6 3.4 -0.2 0.45 0.6 0.2
-0.3 0.55 0.7 0.5 1.2 1.10 0.7 3.8
-0.1 0.89 0.2 0.3 0.5 0.52 0.2 1.6
0.4 0.84 0.3 1.7 0.5 0.48 0.3 1.7
1.3 0.49 -0.3 3.6 0.0 1.02 -0.3 1.0