R Seibert

Helical tomotherapy superficial dose measurements.

Helical tomotherapy is a treatment technique that is delivered from a 6 MV fan beam that traces a helical path while the couch moves linearly into the bore. In order to increase the treatment delivery dose rate, helical tomotherapy systems do not have a flattening filter. As such, the dose distributions near the surface of the patient may be considerably different from other forms of intensity-modulated delivery. The purpose of this study was to measure the dose distributions near the surface for helical tomotherapy plans with a varying separation between the target volume and the surface of an anthropomorphic phantom. A hypothetical planning target volume (PTV) was defined on an anthropomorphic head phantom to simulate a 2.0 Gy per fraction IMRT parotid-sparing head and neck treatment of the upper neck nodes. A total of six target volumes were created with 0, 1, 2, 3, 4, and 5 mm of separation between the surface of the phantom and the outer edge of the PTV. Superficial doses were measured for each of the treatment deliveries using film placed in the head phantom and thermoluminescent dosimeters (TLDs) placed on the phantoms surface underneath an immobilization mask. In the 0 mm test case where the PTV extends to the phantom surface, the mean TLD dose was 1.73 +/- 0.10 Gy (or 86.6 +/- 5.1% of the prescribed dose). The measured superficial dose decreases to 1.23 +/- 0.10 Gy (61.5 +/- 5.1% of the prescribed dose) for a PTV-surface separation of 5 mm. The doses measured by the TLDs indicated that the tomotherapy treatment planning system overestimates superficial doses by 8.9 +/- 3.2%. The radiographic film dose for the 0 mm test case was 1.73 +/- 0.07 Gy, as compared to the calculated dose of 1.78 +/- 0.05 Gy. Given the results of the TLD and film measurements, the superficial calculated doses are overestimated between 3% and 13%. Without the use of bolus, tumor volumes that extend to the surface may be underdosed. As such, it is recommended that bolus be added for these clinical cases. For cases where the target volume is located 1 to 5 mm below the surface, the tumor volume coverage can be achieved with surface doses ranging from 56% to 93% of the prescribed dose.

Out-of-Field Dosimetry Measurements for a Helical Tomotherapy System

Helical tomotherapy is a rotational delivery technique that uses intensity‐modulated fan beams to deliver highly conformal intensity‐modulated radiation therapy (IMRT). The beam‐on time needed to deliver a given prescribed dose can be up to 15 times longer than that needed using conventional treatment delivery. As such, there is concern that this delivery technique has the potential to increase the whole body dose due to increased leakage. The purpose of this work is to directly measure out‐of‐field doses for a clinical tomotherapy system. Peripheral doses were measured in‐phantom using static fields and rotational intensity‐modulated delivery. In‐air scatter and leakage doses were also measured at multiple locations around the treatment room. At 20 cm, the tomotherapy peripheral dose dropped to 0.4% of the prescribed dose. Leakage accounted for 94% of the in‐air dose at distances greater than 60 cm from the machines isocenter. The largest measured dose equivalent rate was 1×10−10 Sv/s in the plane of gantry rotation due to head leakage and primary beam transmission through the systems beam stopper. The dose equivalent rate dropped to 1×10−10 Sv/s at the end of the treatment couch. Even though helical tomotherapy treatment delivery requires beam‐on times that are 5 to 15 times longer than those used by conventional accelerators, the delivery system was designed to maximize shielding for radiation leakage. As such, the peripheral doses are equal to or less than the published peripheral doses for IMRT delivery on other linear accelerators. In addition, the shielding requirements are also similar to conventional linear accelerators. PACS number: 87.53.Dq

Image-guided helical tomotherapy for localized prostate cancer: Technique and initial clinical observations

The purpose of the present study was to implement a technique for daily computed tomography (CT)–based image‐guided radiation therapy and to report observations on treatment planning, imaging, and delivery based on the first 2 years of clinical experience. Patients with previously untreated stage T1 – T3 biopsy‐proven adenocarcinoma of the prostate were considered eligible for treatment with daily CT‐guided helical tomotherapy. The prostate was targeted daily using megavoltage CT (MVCT) images that were fused with treatment‐planning CT images based on anatomic alignments. All patients were treated at 2 Gy per fraction to 76 – 78 Gy (mean: 76.7 Gy). As part of this study, 33 prostate patients were planned, imaged, and treated with a total of 1266 CT‐guided fractions. The prostate, rectum, bladder, femoral heads, and pubis symphysis were visible in one or more slices for all 1266 MVCT image sets. The typical range of measured prostate displacement relative to a 3‐point external laser setup in this study was 2 – 10 mm [3.4 mm standard deviation (SD)] in the anterior–posterior direction, 2 – 8 mm (3.7 mm SD) in the lateral direction, and 1 – 6 mm (2.4 mm SD) in the superior–inferior direction. The obese patients in this study had a substantially larger lateral variation (8.2 mm SD) attributable to mobility of skin marks. The prostate, seminal vesicles, rectum, and bladder anatomy were used to position the patient relative to the desired treatment position without the use of implanted markers. Acute toxicities were within the expected range given the number of patients treated and the dose level. PACS numbers: 87.50.Gi, 87.53.Mr, 87.53.Tf

Verification of helical tomotherapy delivery using autoassociative kernel regression.

Quality assurance (QA) is a topic of major concern in the field of intensity modulated radiation therapy (IMRT). The standard of practice for IMRT is to perform QA testing for individual patients to verify that the dose distribution will be delivered to the patient. The purpose of this study was to develop a new technique that could eventually be used to automatically evaluate helical tomotherapy treatments during delivery using exit detector data. This technique uses an autoassociative kernel regression (AAKR) model to detect errors in tomotherapy delivery. AAKR is a novel nonparametric model that is known to predict a group of correct sensor values when supplied a group of sensor values that is usually corrupted or contains faults such as machine failure. This modeling scheme is especially suited for the problem of monitoring the fluence values found in the exit detector data because it is able to learn the complex detector data relationships. This scheme still applies when detector data are summed over many frames with a low temporal resolution and a variable beam attenuation resulting from patient movement. Delivery sequences from three archived patients (prostate, lung, and head and neck) were used in this study. Each delivery sequence was modified by reducing the opening time for random individual multileaf collimator (MLC) leaves by random amounts. The error and error-free treatments were delivered with different phantoms in the path of the beam. Multiple autoassociative kernel regression (AAKR) models were developed and tested by the investigators using combinations of the stored exit detector data sets from each delivery. The models proved robust and were able to predict the correct or error-free values for a projection, which had a single MLC leaf decrease its opening time by less than 10msec. The model also was able to determine machine output errors. The average uncertainty value for the unfaulted projections ranged from 0.4% to 1.8% of the detector signal. The low model uncertainty indicates that the AAKR model is extremely accurate in its predictions and also suggests that the model may be able to detect errors that cause the fluence to change by less than 2%. However, additional evaluation of the AAKR technique is needed to determine the minimum detectable error threshold from the compressed helical tomotherapy detector data. Further research also needs to explore applying this technique to electronic portal imaging detector data.

TH‐C‐M100E‐02: Optically Stimulated Luminescence of Aluminum Oxide Detectors for Radiation Therapy Quality Assurance

Purpose: The purpose of this experiment was to: 1) Determine if a commercially available Al 2 O 3 detector system used for monitoring personnel exposure could be adapted for use as a radiation therapydosimetry system; and 2) Evaluate the systems performance as an in‐vivo dosimeter and its ability to measure absolute surface dose, isocenter dose, and normal tissues dose in a phantom as part of patient‐specific IMRTquality assurance.Method and Materials: The dosimeters were evaluated for: 1) Signal decay; 2) Field size dependence; 3) Energy dependence; and 4) Angular dependence using the Landauer, InLight MicroStar system. In‐Vivo dosimetry measurements were taken for 22 patients treated on a Varian 21EX. The Landauer system was also tested for its ability to measure absolute dose from helical tomotherapy treatments. Results: The variation between dosimeters was evaluated and found to be ±1.6%. The dosimeters appeared to over‐respond in the first 10 minutes, however, after 10 minutes the chips were within 1 percent of the steady‐state reading. Unlike other detectors, the Al2O3dosimeters showed no field size, energy, or angular dependence. The agreement between the dosimeters and the calculated doses for the in‐vivo dosimetry patients was 2.2±6.1 cGy or 3.7±2.5%. The dosimeters were also tested for their ability to measure absolute dose inside an IMRT phantom. The agreement between the dosimeters and the calculated doses was 0.1±5.3 cGy or 0.7±6.7%. Conclusion: Al 2 O 3 dosimeters can be a convenient, inexpensive alternative to TLDs, MOSFETS, and Diodes. The agreement between calculated and measured doses for in‐vivo dosimetry and IMRT QA is comparable to TLDs, MOSFETS, and Diodes. The dosimeters can be quickly read and analyzed after 10 minutes (to allow time for signal decay). The dosimeters do not appear to have an energy, field size, or angular dependence. In addition, the detectors can be erased and re‐used.

TU‐C‐ValA‐01: Image Fusion and Deformation Using a Genetic Algorithm

Purpose: To develop an image fusion application utilizing genetic algorithms, segment the fused images using treatment‐planning contours, perform a slice‐by‐slice sub‐fusion of the segmented images in order to measure deformation, and then utilize the measured deformation for adaptive therapy on a helical tomotherapy treatment delivery. Methods and Materials: A reference CTimage of a density and spatial resolution phantom was obtained using a MVCT imaging. A secondary MVCT fusion image was obtained with the phantom offset by a known amount with plugs removed or rotated. An image fusion algorithm was created using genetic programming to perform image registration of the MVCT images. Contours of the plugs were used to extract sub‐images that were separately registered and deformed using the genetic algorithm. Adaptive therapy was achieved thorough a treatment deliverysinogram deformation algorithm. The sinogram deformation algorithm was tested using a geometric test case that consisted of a dose triangle with a 5.2‐cm base located inside a dose circle with of 3.14‐cm radius. The test dose pattern was moved by known amounts by deforming the treatment deliverysinogram.Results: The initial genetic fusion of the reference and secondary MVCT images was achieved in approximately 15 generations. The time required to perform the genetic fusion was typically 10 to 15 seconds. The images were fused to within 0.7‐mm of the correct position. At the end of the initial fusion, the genetic algorithm correctly identified one of the plugs as missing in the secondary MVCT dataset. The genetic algorithm correctly segmented the second resolution plug in a sub‐image and deformed it to within 1‐degree and 0.7‐mm of the correct position. The delivery deformation tests moved the dose to within 5‐mm of the desired position. Conclusions: A genetic algorithm has been developed for performing image fusion and simple deformation of defined regions of interest.

WE‐D‐AUD B‐01: Automatic Detection of Delivery Errors Using Autoassociative Kernel Modeling

Purpose: The purpose of this work was to evaluate exit detector data for patients treated with helical tomotherapy. A novel technique was developed for automatically evaluating exit dosimetry using autoassociative non‐parametric modeling, which has the ability to learn complex detector data relationships. Method and Materials: The tomotherapy detector array collects and stores exit dosimetry data during treatment delivery in the form of sinograms, which contain a record of the radiation that exits the MLC and passes through the patient during each treatment. Autoassociative Kernel Modeling (AKM) is a non‐parametric technique that makes parameter estimates by calculating a weighted average of a set of historical data called memory vectors. The memory vectors are contained in what is called a memory matrix, which are used to make predictions. Errors between predicted and test values were calculated using the sum of squared errors, which were used to identify faulty projections within each sinogram.Results: A total of 121 delivery sequences were evaluated from 5 patients (4 Prostates and 1 Head & Neck). Major errors were detected in at least one fraction over the course of treatment for each patient in the study. Other detected errors, while smaller in magnitude, could still be an indication of machine faults. Readings from the ionization chambers located in the head of the accelerator can help classify the type of error, whether they are major anatomical misalignments, MLC positional errors, or machine output errors. Conclusion: The key to the model is determining at what cutoff threshold to set so that all significant errors are detected while keeping reducing the number of false alarms. The results show that the model has the ability to detect errors in the exit dosimetry data. They also suggest that AKM modeling can be a useful tool in monitoring the reliability of radiation delivery.

WE‐E‐AUD C‐01: Prediction of Weight Loss, Tumor Response, and Set‐Up Errors For Head and Neck Patients

Purpose: The objective of this study was to develop a novel tool using Kernel Classification that can be used to automatically identify patients that have, or will have, setup issues requiring intervention such as re‐simulation and/or re‐planning. Method and Materials: Inter‐Fraction motion was retrospectively analyzed for 43 H&N patients that were treated on a helical tomotherapy system. For each patient, CTimages were acquired and transferred to the tomotherapy database for treatment planning and image‐guided patient setup. Both custom aquaplastic masks and a positioning mouthpiece were used in 10 of the 43 patients. Results: Fifteen patients had greater than 10% weight loss during the course of treatment. Six patients had a visible reduction in GTV volume. Immobilization effectiveness decreased as the tumors regressed in size and/or the patients lost weight. If the tumor regression was occurred then time could be scheduled to periodically check the mask fit and to make a new mask if needed. The kernel classification technique correctly identified all 43 H&N patients as either having normal or problematic setup using their respective shift data sets. Classifications were made using only the shift values from the first 14 treatments. The predictive performance seriously degraded when data from fewer than 14 treatments were used. However, adding more did little to improve the performance. Conclusion: This study demonstrated that the kernel regression classification method was able to correctly identify the cause behind IGRT positioning problems for H&N patients. The study validated that IGRT positioning problems cause abnormal problem‐specific distributions in the shift data without using statistical distribution tests. Since this technique is fully automated, it could potentially be used during IGRT sessions to help the therapists decipher the factors that hinder patient setup early in a patients treatment so that the proper precautions can be in place.

TU‐D‐M100F‐02: Automated Quality Assurance for Helical Tomotherapy Using Exit Detector Data

Purpose: The purpose of this work was to develop techniques for utilizing exit detector data in a helical tomotherapy system to automate the QA process. The clinical significance of this study is that an analysis of the exit detector data acquired during treatment delivery could be used to ensure that the correct delivery sequence is being administered to the patient. Method and Materials:Software applications were developed to automatically analyze uncompressed detector data. To test this software, a MLC test sequence was designed that allows each MLC leaf to be tested using the exit detector data. The goals of this test are 1.) To identify MLC problems (stuck leafs, bad valves, etc…) before failure, and 2.) Perform QA Tests on MLC parameters that affect the delivered dose. Additionally, a clinical test case was created using the treatment delivery sequences for a head & neck patient. The original treatment delivery sequence was modified and 12 known MLC errors were inserted in the MLC controller file. The procedures were delivered and the software was used to analyze the exit detector data. Results: The MLC QA Test was delivered on four occasions with two MLCs. A software application developed by the investigators was then used to analyze the exit detector data from these deliveries. The software correctly performed Latency Tests, Projection Centering Tests, and MLC Transit Time Tests. For each delivery, the tests showed that the MLCs were properly functioning. For the Head & Neck test case, the shape‐detection algorithm was able to identify 11 out of 12 known MLC errors >10 msec. Conclusion: A technique was developed for performing automated QA of the MLC and individual patient deliveries using exit detector data in a helical tomotherapy system. With the tomotherapy detector array, errors in MLC position > 10 msec have been detected.

SU‐GG‐AUD‐01: Exit Dosimetry Treatment Verification Using Auto‐Associative Kernel Regression

Purpose: The purpose of this work was to develop a novel technique for automatically evaluating exit dosimetry on tomotherapy systems using auto‐associative modeling that is robust and has the capability to learn complex detector data relationships, even with detector data with a low temporal resolution and beam attenuation from the patient. Method and Materials: Delivery sequences from 3 patients were used in this study. Each delivery sequence was modified by reducing the opening time for random individual MLC leaves by random amounts. The error and error‐free treatments were delivered with different phantoms in the path of the beam. Multiple auto‐associative kernel regression (AAKR) models were developed and tested by the investigators using combinations of the sinogramdata sets.AAKR is a non‐parametric model that is used to predict correct values when supplied a group of sensor values that is corrupted. Models were tested using the data containing errors. However, models were never developed with data which had the same object in the path of the beam as the dataset it was testing. This allowed the testing of the models error detection capabilities in the presence of attenuation. Results: The results show that the model correctly distinguished the MLC positional error from changes in attenuation. The model identified errors in compressed detector data that had been summed over 94 frames. Generally, errors greater than 7 milliseconds were visually discernable. Some smaller errors could be detected, but it depended on the position of the erroneous leaf in the projection and the actual projection shape. Conclusion: The results presented suggest that AAKR modeling could be used to monitor and eventually improve the reliability of radiation delivery. This method has the potential to play a noteworthy role in determining and possibly correcting for the types of machine‐related errors that occur during actual patient treatments.