Upstate New York Association of Physicists in Medicine, Inc. and Western New York Chapter of the Health Physics Society Joint Meeting
(A Chapter of the AAPM)
Spring Meeting – Friday May 15, 2009
Whipple Auditorium, Univ. of Rochester Medical Center

Room 2-64-24, 601 Elmwood Ave, Rochester, NY 14642

 

 

12:00 pm – 5:30 pm    Sponsoring Vendor Exhibits:

 

Vision RT                                                                              North American Scientific

Impac Medical Systems                                                      Upstate Linac Services, LLC

Core Oncology                                                                 LACO, Inc

IBA Dosimetry

                                               

10:30 AM-11:30 AM    Business Meeting

12:00              Lunch   Sponsored by      Upstate Linac Services, LLC

 

 

12:30              Refreshment and Vendor Exhibit

 

 

1:00

Meeting Introduction

Walter O’Dell

UNYAPM President

Proffered Paper Session (8 minute talks, 2 min QA)

 

1:10

Random walk model for predicting patterns of microscopic glioma spread using DTI: A prospective study

A. Krishnan, D. Davis, P. Okunieff, and W. O'Dell

1:20

 

Fetal Dose Reduction in Different Shielding Scenarios During Thoracic CT

K. Greene-Donnelly, K Ogden, M. Roskopf

1:30

Extracranial Dose measurements for the Leksell Gamma Knife Model 4C using Gafchromic EBT Film

T. Tran, C.D. Arndt, J.P. Steinman, and M.B. Podgorsak

1:40

An Optical Guidance Technique for Patient Position during Breast Radiotherapy

J Schmitt, K Hoffmann, M Bakhtiari, D Nazareth , H Malhotra

 

1:50

Radiographs In Pretreatment IMRT QA: Are Films Necessary In Addition To Electronic Quality Assurance Methods?

D W Bailey, S F de Boer and M B Podgorsak

2:00

Effect of surface waves on the dosimetric measurements in water tanks

M. Bakhtiari, S. de Boer, and M. B. Podgorsak

2:10-2:30

 Development of an Adequate Quality Assurance Program

for Gamma Probes

Rich Harvey

2:30

Refreshment , Vendor Exhibits and Poster Viewing* – Sponsored by

                                                  Upstate Linac services

3:15

Invited Speaker

Will They Ever Learn? The Public Education Game

 Howard Dickson

     President-elect of the national

     Health Physics Society

4:20

Unaccounted Intracranial Dose during Patient Repositioning with the Gamma Knife APS Device

 

T. Tran, T.R. Stanley, H.K. Malhotra, S.F. deBoer, D. Prasad, M.B. Podgorsak,

 

4:30

A comprehensive quality assurance procedure for ultrasound-guided radiation therapy

Dinko Plenkovich, Matthew B. Podgorsak, Jubei Liu

4:40

Dose Perturbation from Implanted I-125 Seeds in External Beam Therapy for Prostate Cancer

JP Steinman, M Bakhtiari, DP Nazareth, HK Malhotra

4:50

Respiratory gating using online automatic segmentation of pulmonary nodules in megavoltage

electronic portal images using a level set method

JS Schildkraut, J Gomez, A Singh, D Nazareth, HK Malhotra

 

5:00

Invitation for  a tour of the new URMC cancer center

Dr. Schell

                                                                                                             

     ADJOURN

 

 

Driving Directions to the University of Rochester Medical Center

 

From the West:  New York State Thruway to Exit 47.

  -  Take exit #47/LEROY (RT-19)/ROCHESTER onto I-490 E (Toll applies) - go 19.63 mi

  -  Take exit #9B/AIRPORT onto I-390 S - go 2.9

  -  Take exit #17/SCOTTSVILLE RD. - go 0.2 mi

  -  Turn Left on SCOTTSVILLE RD(RT-383) - go 0.6 mi

  -  Bear Right on ELMWOOD AVE - go 0.9 mi

Parking is available in the Ramp Garage.

 

From Rochester Airport (ROC):

  -  Take exit #9B/AIRPORT onto I-390 S - go 2.9

  -  Take exit #17/SCOTTSVILLE RD. - go 0.2 mi

  -  Turn Left on SCOTTSVILLE RD(RT-383) - go 0.6 mi

  -  Bear Right on ELMWOOD AVE - go 0.9 mi

 

From the East: New York State Thruway to Exit 46.

  -  Take exit #46/ROCHESTER/CORNING onto I-390 N toward ROCHESTER (Toll applies) - go 6.9mi

  -  Take exit #16/E HENRIETTA RD/W HENRIETTA RD (RT-15) - go 0.2 mi

  -  Turn Right on E HENRIETTA RD(RT-15A) - go 0.9 mi

  -  Bear Right on MT HOPE AVE(RT-15) - go 0.2 mi

  -  Turn Left on ELMWOOD AVE - go 0.3 mi

 

From the South:  390 North and follow the directions above when coming from the East.

 

  

 

 

 

 

 

 

 

 

                                              

 

UNYAPM SPRING MEETING PROCEEDINGS

University of Rochester Medical Center, Rochester, NY

May 15, 2009

 

Unaccounted Intracranial Dose during Patient Repositioning with the Gamma Knife APS Device

 

T. Tran, T.R. Stanley, H.K. Malhotra, S.F. deBoer, D. Prasad, M.B. Podgorsak,

Roswell Park Cancer Inst., Buffalo NY

 

Purpose: Measure unaccounted dose delivered to the target site and its periphery from the defocus and transit beam during automatic positioning system (APS) repositioning for Gamma Knife Radiosurgery.  

Methods and Materials: A stereotactic head-frame was attached to a 16cm diameter spherical phantom with a calibrated ion-chamber at its center.  Using a fiducial-box to determine the coordinates of the target, a CT scan with 1mm slice thickness was taken of the phantom and registered in the GammaPlan TPS.  10Gy to the 50% isodose line was prescribed to the target site for all measurements.  Plans were generated for the 18mm, 14mm and 8mm helmets with varying number repositions for each plan to determine the relationship of measured dose with number of repositions of the APS system and helmet size.  The shot isocenter was identical in the entire study and there was no movement of APS between various shots; this allows for measurement of transit dose (couch moves from the focus to defocus position and back) and the least defocus dose (at defocus couch position).  The couch was suspended in the defocus position allowing intracranial defocus dose measurements. 

Results:  Dose increases with frequency of repositioning and collimator size. Overdose of up to 5.71±0.07% at target position can result from couch transfer.  Dose rate of 8.81±0.41cGy/min (18mm helmet) and 5.89±0.51cGy/min (8mm helmet) where measured.  During couch transit, the target receives more dose than peripheral regions; in the defocus position, the greatest dose is superior on the phantom where dose rate is 4.91±0.01cGy/min. 

Conclusion: APS repositioning results in additional dose to the target site and its periphery for multi-shot runs.  Doses in superior regions should be monitored due to epilation.  Consideration of conformity index is suggested when generating a treatment plan as risk of toxicity is a concern, especially around critical structures (such as the optic nerve).  Application of a timer error would account for these doses during couch transit and APS repositioning; this would also improve the accuracy of the prescription dose.

 

 

Fetal Dose Reduction in Different Shielding Scenarios During Thoracic CT

 

K. Greene-Donnelly, K Ogden, M. Roskopf, 
Upstate Medical University, Syracuse NY
 

Purpose:  To determine the potential for reducing fetal dose in early pregnancy during thoracic computed tomography by shielding the patient’s abdomen and pelvis.

Method and Materials:  An anthropomorphic phantom (Rando) representing a medium sized adult was used to measure relative tissue doses in the abdomen/pelvis during thoracic CT scanning.  TLD’s were used to measure the dose along the central axis of the phantom from the level of the adrenals to the location of the uterus.  The phantom was scanned on a GE Lightspeed VCT 64 slice scanner using 120 kVp, 750 mAs, and pitch of 0.984 to increase the TLD signal.  Scans were performed with no shielding, with shielding (lead apron) on the anterior aspect of the phantom over the abdomen and pelvis, and with shielding on both the anterior and posterior aspects of the phantom. Tissue doses were normalized to the in-scan value measured at the level of the adrenals.

Results:  Tissue doses decreased exponentially with increasing distance from the bottom of the scanned anatomy.  The rates of decay were -0.18 ± 0.010 cm-1, -0.21 ± 0.0024 cm-1, and -0.23 ± 0.0018 cm-1 for the no shielding, half shielding, and full shielding cases, respectively.  The dose values at the level of the uterus were 0.99%, 0.65%, and 0.51% of the dose at the level of the adrenals, for the no shielding, half shielding, and full shielding cases, respectively.

Conclusion:  These data show that the total absolute dose received by a fetus early in the pregnancy may be reduce by approximately 1/2 during thoracic CT by use of shielding on the abdomen/pelvis.  For a clinical technique of 120 kVp, 150 mAs, and pitch of 1.375, this would reduce the fetal dose from an estimated 0.1 mGy to 0.05 mGy.

 

 

Extracranial Dose measurements for the Leksell Gamma Knife Model 4C using Gafchromic EBT Film

 

T. Tran, C.D. Arndt, J.P. Steinman, and M.B. Podgorsak,

Roswell Park Cancer Institute, Buffalo, NY

 

Objective: To obtain measurements of scatter dose using the Rando phantom and Gafchromic film in critical, extracranial organs for patients undergoing Gamma Knife Radiosurgery. 

Method and Materials: A stereotactic frame was attached to the head of an anthropomorphic Rando phantom.  Using a fiducial box to determine the coordinates of the target (center of the Rando head), a CT scan was taken and registered in the GammaPlan treatment planning system where a dose prescription of 25Gy to the 50% isodose line was applied to the target site.  The plan was generated for the 18 mm collimator size helmet with a single shot run using the automatic positioning system (APS).  Multiple (2 to 5) 2”x2” Gafchromic EBT films were placed between each slice of the Rando body phantom, from the neck to pelvic region (phantom slice 8 to 31), and analysis was done using a Vidar VXR-16 scanner with the RIT113 Version5.1 analysis software.  An H&D curve was created using Gafchromic EBT film, a calibrated 0.05cm3 ion chamber, the Keithley 35617EBS Programmable Dosimeter, and a framed 16cm diameter sphere phantom (with film and ion chamber inserts). 

Results: The dose was 11.3±1.1 cGy to the thyroid, 8.7±1.3 cGy to the thymus, 5.1±0.4 cGy average dose to the adrenal glands, 1.0±0.2 cGy to the ovaries, 0.8±0.3 cGy for the testes, 3.3±0.4 cGy to the pancreas and 2.7±0.5 cGy to the colon.

Conclusion: Extracranial doses depended on total target dose and the distance the organ was from isocenter during treatment.  For all organs, greater prescription doses lead to greater doses to extracranial organs.  Dose to the organs also decreased with increasing distance from the focal point.  The extracranial doses are well within tissue tolerances and are comparable with other studies.  Future projects will include comparisons with the Leksell Gamma Knife PERFEXION.  Doses are low but may be considered for younger patients with longer life expectancy.

 

 

An Optical Guidance Technique for Patient Position during Breast Radiotherapy

 

J Schmitta, K Hoffmannb, M Bakhtiaria, D Nazaretha , H Malhotraa

a.) Roswell Park Cancer Institute b.) Toshiba Stroke Center, University at Buffalo

 

Purpose: Breast radiotherapy, particularly IMRT, involves large dose gradients and difficult patient positioning problems.  A critical requirement for successful treatment is accurate reproduction of the patient’s position assumed during CT simulation and planning.  Solving this problem using a simple optical system requires careful imaging geometry calibration.  We have developed an optical image-guided technique, which assists in accurately and reproducibly positioning the patient, by displaying her real-time optical image superimposed on a perspective projection image of her 3D CT data.

Methods. The Single Projection Technique (SPT) accurately determines the 3-D position and orientation of a camera from a single image acquired of a known model.  A calibration jig, composed of ten identifiable reflecting spheres, was constructed and CT imaged to provide this model. The coordinates of each point were determined with respect to a fiducial marker.  To implement our method, a digital photograph of the jig is acquired, and a centroid-finding technique is applied to this image.  The two-dimensional coordinates of each sphere, along with its 3D coordinates serves as input to the SPT program, which calculates the coordinates and orientation of the camera.  Using this information, 3D CT patient data is projected onto the camera’s imaging plane, and is displayed on a monitor, superimposed on the real-time patient image. This enables the therapist to view both the patient’s current and desired positions, and guide proper patient positioning. 

Results: The SPT can determine the position and orientation of the camera to an accuracy of 0.2 cm and 0.3°, respectively.  Investigations are ongoing to determine the accuracy and reproducibility of our method, based on film measurements performed on a breast phantom.

Conclusion: We have developed a method to calibrate an optical camera system and superimpose a perspective projection of a CT image on a patient’s real-time optical image.  Displaying this visual information will assist in accurate setup during breast radiotherapy.  Future work will enable us to quantify the setup and dose delivery accuracy of this technique.

 

 

Radiographs In Pretreatment IMRT QA: Are Films Necessary In Addition To

Electronic Quality Assurance Methods?

 

D W Bailey, S F de Boer and M B Podgorsak,  

Roswell Park Cancer Institute, Buffalo, NY 14263

 

Purpose: The use of radiographs for IMRT QA has several disadvantages, including time-consuming and resource-demanding development, scanning, and dose calibration.  Although radiographs offer the best resolution in measurement of dose distributions, electronic QA methods, e.g. diode arrays and electronic portal imaging devices (EPIDs), have become standard for efficiently and accurately verifying dose distributions.  There is still a question as to whether or not radiographs are necessary in addition to electronic QA to qualitatively verify the geometric accuracy of delivered fluences.  However, due to the finite ability of the human eye to compare film exposures to planning system printouts, electronic QA methods may detect geometric inaccuracies before the same error is recognizable on film.   This study addresses the question of whether or not the qualitative use of films contributes significantly to the pretreatment verification process. 

Methods: An IMRT fluence was delivered on film, and errors were systematically introduced by omitting portions of the fluence, decreasing the number of delivered monitor units and control points proportionally. This method simulates a communication error between planning and delivery systems.  The same modified fluence was delivered on MapCHECK (Sun Nuclear Corporation, Melbourne FL) and compared to the TPS verification plan using gamma evaluation of 3%, 3 mm. The process was repeated, omitting increasingly greater portions of the fluence until the IMRT QA failed, either by errors observed on the film or by MapCHECK gamma analysis below 85% passing points.  This procedure was repeated on four fluences with increasing levels of modulation. 

Results: For every modified fluence that failed IMRT QA, the errors were apparent in MapCHECK before they were observed on film.  Even with as much as 20% of control points omitted, the radiographs for all fluences appear virtually unchanged to the naked eye.  Contrastingly, MapCHECK analysis of the same fields failed due to omission of control points by as little as 6.3% (average) from the center of the fluences, and 19.3% (average) from the outer portions of the fluences.    

Conclusions: If a portion of an IMRT fluence is omitted due to data transfer errors, qualitative analysis of radiographs does not enhance the ability to detect such errors during pretreatment IMRT QA if a diode array with acceptable resolution is utilized.  However, because such errors may not be detected by 3%, 3mm gamma evaluation until 10-20% of control points are lost, gamma analysis of IMRT fields should always be accompanied by comparison of the isodose distributions in measured and predicted fluences.  Further experimentation must be conducted to determine if MapCHECK eliminates the need for radiographs in the event of other types of errors. 

 

 

Effect of surface waves on the dosimetric measurements in water tanks 

 

M. Bakhtiari, S. de Boer, and M. B. Podgorsak,

 Roswell Park Cancer Institute, Buffalo, NY 14263.

 

Purpose: To study the effect of surface water waves on the accuracy of ionization measurements in large scanning water phantom.  

Methods and Materials: Profile measurements were taken in a PTW water tank (50cm×50cm×50cm) filled with water. The detector (ion-chamber) is attached to a variable speed movable arm that moves in a Cartesian coordinate system. The arms speed was varied from 1 mm/s to 50 mm/s, the dose collection time was 0.3 ms, and the spatial resolution was selected to be 1 mm. Profiles were measured at a depth of 12.6 mm (R50) for a 10cm×10cm, 4 MeV electron beam. Two sets of experiments were carried out; 1) after each profile measurement the arm was left at the end point. For starting the new profile the arm had to come back to the starting point and immediately start the new profile, 2) after each measurement the arm was immediately brought back to the starting point and the next profile measurements were started 2 minutes later.

Results:  The amplitude of the surface water waves increases with increasing the speed of arm. Consequently some errors in the measurements were observed. The measurements with slow moving arms (1 mm/s) were more reproducible and demonstrated less fluctuation. The reproducibility decreased and fluctuations increased with increasing the speed. When the measurements started a while after the arm was brought to starting point the accuracy increased, otherwise even with slow moving arms some errors were observed in the measurements.

Conclusion:  The moving arms in large water tanks can have an impact in dosimetry.  It was found the surface waves can cause errors of 3% and 8% for slow moving and fast moving arms, respectively. 

 

 

Development of an Adequate Quality Assurance Program for Gamma Probes

 

Richard P. Harvey, DrPH, ABSNM, CHP, CMLSO. CLSO, LMP

Director of Radiation Safety and Radiation Safety Officer, Roswell Park Cancer Institute

 

Gamma probes are a radiation detection tool used in surgery to identify tissue for resection and their oncologic applications have become fairly common.  Many surgeons and surgical departments purchase these devices for use regardless of radiation physics assessment and thought for Quality Assurance.  Many physicians believe the limited quality control recommended by the sales representative is enough to provide adequate performance evaluation of gamma probes. 

      Each licensee must develop an adequate Quality Assurance Program to ensure proper function and adequate calibration of instruments used for patient care.  These methods need to be established and communicated among licensees for the benefit of surgical patients treated at all healthcare facilities. 

 

Random walk model for predicting patterns of microscopic glioma spread using DTI: A prospective study

 

A. Krishnan, D. Davis, P. Okunieff, and W. O'Dell

Depts of Biomedical Engineering and Radiation Oncology, University of Rochester, Rochester, NY

 

Purpose: The current methods of determining treatment margins needed to encompass microscopic tumor spread for Stereotactic Radiotherapy (SRT) are often inadequate as recurrences/secondary tumors often occur at the boundary of the treatment margin. We hypothesize that paths of elevated water diffusion along the white matter tracts provide a preferred path for migration of glioma cells. If our hypothesis is true, then future SRT plans would be modified to provide elongated margins along white matter tracts from the primary tumor, thereby targeting tissue with unseen, microscopic spread of tumor cells and hence reducing the incidence of recurrence/secondary tumors. We present here the pattern of glioma spread observed in follow-up MR images and compare it with the results of anticipated tumor spread from our predictive random walk model of cell migration based on DTI obtained prospectively.

Methods and Materials: We acquired high resolution DTI datasets of glioma patients in a prospective study to validate the predictive power of our hypothesis. As per standard of care the primary tumor was surgically resected followed by SRT. For our protocol the patients were then imaged either pre-surgically or post-surgically before SRT after the reduction of edema.  Three volunteers and thirteen patients with gliomas were imaged. Following SRT, patients were given repeated clinical MRI follow-ups at regular intervals to identify early incidence of tumor recurrence. Our method involved DTI acquisition and processing, followed by the application of a constrained random walk model for cell migration.  1) The DTI datasets were reconstructed with Camino/DTIStudio and PDD and Fractional Anisotropy (FA) were obtained. 2) The migration of each cell from the surface voxel was simulated independently. The uncertainty in the direction of cell migration about the PDD was determined based on the FA value of the voxel. The PDD was given by the in-plane and out-of-plane solid angles. The uncertainty in cell migration was ±35°, ±20° and ±10° about the PDD when the FA was 0-0.3, 0.3-0.6 and 0.6-1, respectively. 3) At each step the direction of migration was decided randomly within the uncertainty range. 4) When the cell was on the tumor surface it was constrained to move away from the center of the tumor. 5) The probability of cell migration was defined as the number of cells found in or passing through each voxel after a fixed number of steps.

Results and Conclusions: Of the 13 patients recruited to date, six have had recurrence/secondary tumors. Two patients had secondary tumors outside the treatment margin and in both of these patients there was a high correlation between the areas of high cell concentration predicted by our random walk model and the location of secondary tumors (Figure 1). Four patients had recurrences within the treatment margin. In one of these patients the areas of high cell concentration from the random walk model predicted the direction of tumor spread (Figure 2). For recurrences outside the treatment margin our hypothesis appears to be valid.

 

 

A comprehensive quality assurance procedure for ultrasound-guided radiation therapy

Dinko Plenkovich, Roswell Park Cancer Institute, WCA Cancer Treatment Center, Jamestown, NY

Matthew B. Podgorsak, Jubei Liu, Roswell Park Cancer Institute, Buffalo, NY

Purpose:  Develop a method for evaluation of the entire process for ultrasound guidance in the targeting of cancer treatment sites.  The vendor has provided only the phantoms for: a) camera verification by registering the ultrasound coordinate system to that of the linac room coordinate system, and b) ultrasound probe verification, which informs the ultrasound system of the condition of the ultrasound probe, including the integrity of ultrasound imaging and the stability of the optical tracking array relative to the probe. 

Method and Materials:  An ultrasound phantom was developed and scanned on the CT scanner.  The images were transferred to the treatment planning system and the structures in the phantom were contoured.  The treatment plan was exported to an ultrasound system for positioning an anatomical target to the linac isocenter for extracranial radiation therapy 

Results:  By tracking the probe’s position and matching pretreatment isocenter CT image contours to image models, structure position variances were determined and corrected by repositioning the target. 

Conclusion:  The phantoms provided by the vendor are not sufficient.  In one of our affiliate institutions, the vendor-recommended morning quality assurance method failed to detect a 2-cm difference between the actual and perceived position of the prostate.  It is necessary to evaluate the whole process from the CT scan to the comparison of the ultrasound images with the contours created in the treatment planning system.  An appropriate phantom should be available for this evaluation.

 

 

 

Dose Perturbation from Implanted I-125 Seeds in External Beam Therapy for Prostate Cancer

 

JP Steinman, M Bakhtiari, DP Nazareth, HK Malhotra,

Roswell Park Cancer Institute, Buffalo, NY

 

Purpose:  Many times a suboptimal dose distribution resulting from I-125 seeds in prostate brachytherapy is salvaged by giving additional radiation dose using 3DCRT/IMRT.  In standard treatment planning, the dosimetric perturbations introduced by the existing seeds are usually ignored.  Present study aims at studying these perturbations for 6MV and 18MV beams within a phantom setting in region immediately behind the seed.

Methods and Materials:  Three Kodak X-OmatV films were placed on top of 10cm of Solid Water at 100cm SSD.  On top of the films a single non-radioactive (preactivated) seed was placed and aligned parallel in the longitudinal direction under 1cm bolus and 4cm Solid Water for a total buildup of 5cm.  A 1cm x 1cm field was setup and irradiated with 10MU of 6MV and 18MV photons.  A second set of measurements was obtained using three seeds each separated vertically by 0.5cm bolus material allowing the study of the interseed shielding effect.  Control fields were irradiated with no seeds.  All the films belonged to the same batch and were processed simultaneously.  The films were scanned using a Vidar VXR-16 scanner and analyzed using RIT 113 Version 5.1 obtaining profiles in the transverse and longitudinal direction.   Additional external beam treatment plans were generated for a prostate patient implanted with I-125 seeds using a Monte Carlo software [VMC++] with and without accounting for the effect of I-125 seeds for both 6 MV and 23MV.

Results:  For the single seed measurement, at about 0.5mm from the seed (top film), the maximum change in dose from having no seed was 27.1% (6MV) and 13.4% (18MV).  The three seed measurement revealed 24.1% (6MV) and 11.1% (18MV).  Results of external beam planning using Monte Carlo simulation (VMC++) carried out on a patient implanted with I-125 seeds will be presented.

Conclusion:  The dose perturbation caused by the I-125 seeds is significant locally around the seed.  This can be seen by the fact that the change in the dose profile is independent of the number of seeds spaced intermitted above the seed.  

 

 

Respiratory gating using online automatic segmentation of pulmonary nodules in megavoltage

electronic portal images using a level set method

 

JS Schildkraut, J Gomez, A Singh, D Nazareth, HK Malhotra,

 Carestream Health, Inc. Rochester NY, SUNY Buffalo

 

Purpose:  Tumors of the thoracic region present unique problems during their treatment due to the associated motion of the tumor.  Various methods like respiratory gating alone or in conjunction with abdominal compression have been developed to either reduce the motion or to gate the treatment.  Unfortunately, conventional respiratory gating relies on an external surrogate.  Studies have shown the inadequacy of this approach in many cases.  Fortunately, majority of the lung tumors are surrounded by low density lung tissues allowing an easier identification even with megavoltage imaging.  The present work focus on the development of an alternative system which harnesses the density differences between the tumor and its surroundings lung tissue directly, thereby, removing the necessity of an external surrogate system. 

Methods and Materials:  A total of 7 patients of non-small cell lung cancer were used in this study.  The study was carried out on a Varian Trilogy unit equipped with electronic portal imaging system [EPID] employing an AS-1000 system.  The system has a resolution of 1024x1024 and provides an active area of 30x40 cm2.   During the treatment, a custom designed image acquisition template was applied which captured the images at every 10% of the delivered monitor units.  The resultant images were exported in dicom format.  The treatment plan along with the raw scan data and structure was also exported in DIOCM format.

Results:  The location of the pulmonary nodule is delineated in each CT slice in which it appears. A digitally reconstructed radiograph (DRR) is calculated from the treatment planning CT scan using the source and detector geometry of the EPID. In the process of calculating the DRR, the projection of the nodule in the DRR is determined. The nodule’s projection in the DRR is subsequently used as a shape prior. Nodules are segmented in each portal image using a level set segmentation algorithm which includes an energy term that is minimized when the shape of the segmented region matches the shape prior. The nodule segmentation method was tested on a series of 20 portal images of a nodule that is located just above the diaphragm. The nodule has an average distance from the center of the portal of 5.68 mm. Due to respiratory motion, the distance between the nodule and the portal center has a standard deviation, minimum, and maximum distance of 1.79, 2.44, and 9.13 mm, respectively. The distance between the nodule and segmented region center was also measured. The average distance is 1.98 mm and the standard deviation, minimum, and maximum distance is 0.96, 0.13, and 4.12 mm, respectively. These preliminary results suggest that if the radiation treatment system were to use the results of the nodule segmentation algorithm to track the nodule, the average position error between the portal and nodule center could be reduced by 35%. Also, the maximum position error can be reduced by 45%. The segmentation algorithm has a runtime of 4 seconds on a 2.33 GHz Intel® Core™2 Duo Desktop Processor E6550 with 3.48 GB of RAM.

Conclusion:  The purposed method utilizes the tumor motion directly, thereby, eliminating an external surrogate system and its associated inaccuracies.  Efforts are in progress to harness the massive parallel computational power of the new breed of graphic cards [GPUs] which will enable the execution of the algorithm in real-time are in progress.   This method can then control linac beam on in standard RPM based systems. The method can be used for both amplitude based as well as phase based respiratory gating techniques including breath-hold technique of treatment.