-
[show abstract]
[hide abstract]
ABSTRACT: During the last decade, studies have shown that 3D list-mode ordered-subset expectation-maximization (LM-OSEM) algorithms for positron emission tomography (PET) reconstruction could be effectively computed and considerably accelerated by graphics processing unit (GPU) devices. However, most of these studies rely on pre-calculated sensitivity matrices. In many cases, the time required to compute this matrix can be longer than the reconstruction time itself. In fact, the relatively long time required for the calculation of the patient-specific sensitivity matrix is considered as one of the main obstacle in introducing a list-mode PET reconstruction algorithm for routine clinical use. The objective of this work is to accelerate a fully 3D LM-OSEM algorithm, including the calculation of the sensitivity matrix that accounts for the patient-specific attenuation and normalization corrections. For this purpose, sensitivity matrix calculations and list-mode OSEM reconstructions were implemented on GPUs, using the geometry of a commercial PET system. The system matrices were built on-the-fly by using an approach with multiple rays per detector pair. The reconstructions were performed for a volume of 188×188×57 voxels of 2×2×3.15 mm(3) and for another volume of 144×144×57 voxels of 4×4×3.15 mm(3). The time to compute the sensitivity matrix for the 188×188×57 array was 9 s while the LM-OSEM algorithm performed at a rate of 1.1 millions of events per second. For the 144×144×57 array, the respective numbers are 8 s for the sensitivity matrix and 0.8 million of events per second for the LM-OSEM step. This work lets envision fast reconstructions for advanced PET applications such as real time dynamic studies and parametric image reconstructions.
Physics in Medicine and Biology 09/2012; 57(19):6279-93. · 2.83 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: To determine the extent of gold fiducial marker (FM) migration in patients treated for prostate cancer with concurrent androgen deprivation and external-beam radiation therapy (EBRT).
Three or 4 gold FMs were implanted in 37 patients with prostate adenocarcinoma receiving androgen deprivation therapy (ADT) in conjunction with 70-78 Gy. Androgen deprivation therapy was started a median of 3.9 months before EBRT (range, 0.3-12.5 months). To establish the extent of FM migration, the distance between each FM was calculated for 5-8 treatments once per week throughout the EBRT course. For each treatment, the distance between FMs was compared with the distance from the digitally reconstructed radiographs generated from the planning CT. A total of 281 treatments were analyzed.
The average daily migration was 0.8 ± 0.3 mm, with distances ranging from 0.2 mm-2.6 mm. Two of the 281 assessed treatments (0.7%) showed migrations >2 mm. No correlation between FM migration and patient weight or time delay between ADT and start of EBRT was found. There was no correlation between the extent of FM migration and prostate volume.
This is the largest report of implanted FM migration in patients receiving concomitant ADT. Only 0.7% of the 281 treatments studied had significant marker migrations (>2 mm) throughout the course of EBRT. Consequently, the use of implanted FMs in these patients enables accurate monitoring of prostate gland position during treatment.
International journal of radiation oncology, biology, physics 04/2012; 84(1):e7-e12. · 4.59 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: To validate GPUMCD, a new package for fast Monte Carlo dose calculations based on the GPU (graphics processing unit), as a tool for low-energy single seed brachytherapy dosimetry for specific seed models. As the currently accepted method of dose calculation in low-energy brachytherapy computations relies on severe approximations, a Monte Carlo based approach would result in more accurate dose calculations, taking in to consideration the patient anatomy as well as interseed attenuation. The first step is to evaluate the capability of GPUMCD to reproduce low-energy, single source, brachytherapy calculations which could ultimately result in fast and accurate, Monte Carlo based, brachytherapy dose calculations for routine planning.
A mixed geometry engine was integrated to GPUMCD capable of handling parametric as well as voxelized geometries. In order to evaluate GPUMCD for brachytherapy calculations, several dosimetry parameters were computed and compared to values found in the literature. These parameters, defined by the AAPM Task-Group No. 43, are the radial dose function, the 2D anisotropy function, and the dose rate constant. These three parameters were computed for two different brachytherapy sources: the Amersham OncoSeed 6711 and the Imagyn IsoStar IS-12501.
GPUMCD was shown to yield dosimetric parameters similar to those found in the literature. It reproduces radial dose functions to within 1.25% for both sources in the 0.5< r <10 cm range. The 2D anisotropy function was found to be within 3% at r =5 cm and within 4% at r = 1 cm. The dose rate constants obtained were within the range of other values reported in the literature.
GPUMCD was shown to be able to reproduce various TG-43 parameters for two different low-energy brachytherapy sources found in the literature. The next step is to test GPUMCD as a fast clinical Monte Carlo brachytherapy dose calculations with multiple seeds and patient geometry, potentially providing more accurate results than the TG-43 formalism while being much faster than calculations using general purpose Monte Carlo codes.
Medical Physics 07/2011; 38(7):4101-7. · 2.83 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: The impact of tissue heterogeneity and interseed attenuation is studied in post-implant evaluation of five clinical permanent breast (103)Pd seed implants using the Monte Carlo (MC) dose calculation method. Dose metrics for the target (PTV) as well as an organ at risk (skin) are used to visualize the differences between a TG43-like MC method and more accurate MC methods capable of considering the breast tissue heterogeneity as well as the interseed attenuation. PTV dose is reduced when using a breast tissue model instead of water in MC calculations while the dose to the skin is increased. Furthermore, we investigate the effect of varying the glandular/adipose proportion of the breast tissue on dose distributions. The dose to the PTV (skin) decreases (increases) with the increasing adipose proportion inside the breast. In a complete geometry and compared to a TG43-like situation, the average PTV D(90) reduction varies from 3.9% in a glandular breast to 35.5% when the breast consists entirely of adipose. The skin D(10) increases by 28.2% in an entirely adipose breast. The results of this work show the importance of an accurate and patient-dependent breast tissue model to be used in the dosimetry for this kind of low energy implant.
Physics in Medicine and Biology 08/2010; 55(16):4547-61. · 2.83 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: To assess the influence of fiducial marker (FM) migration on the matching quality in external beam radiation therapy (EBRT) for prostate cancer.
The position of FMs were identified using on-board kV imaging (OBI) and their 3-D position established using an in-house reconstruction algorithm for 31 patients with prostate adenocarcinoma. To carry out the match, the positions were overlaid on the digitally reconstructed radiographs (DRR) generated from the planning CT. The distance between each FM was calculated for seven treatments throughout the EBRT course. Four radiotherapy technologists were asked to independently perform and rate the match from OBI to DRR which was then correlated to the extent of FM migration.
All the matches were rated by at least three radiotherapy technologists as "very easy" ("easy" subgroup) for 24 patients (77%), while the other seven patients had their match rated less than "very easy" and considered the "not easy" subgroup. The average daily FM migration was 0.93+/-0.34 mm for the "easy" subgroup vs. 1.82+/-0.75 mm for the latter. An average migration >2 mm was seen in five/seven patients in the "not easy" subgroup as compared to none in the "easy" subgroup. There was a trend towards less FM migration and better matching if the planning CT was done later than the day of the FM implant (p=0.093).
FM migration >2 mm predicts for a more difficult matching process; PTV margins might have to be adjusted or the planning CT repeated.
Radiotherapy and Oncology 04/2010; 96(1):43-7. · 5.58 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: To compare low (mean 0.44, SD ± 0.0163 mCi) with high source activity (0.61 ± 0.0178 mCi) in I(125) permanent seed brachytherapy regarding seed loss, dosimetric outcome, and toxicity.
The study included 199 patients with prostate cancer treated by permanent seed brachytherapy alone: the first 105 with seeds of lower activity (first cohort), the following 94 with higher seed activity (second cohort). The V100, V150, V200, and D90 were analyzed on the CT scan 30 days after implantation (CTD30). The V100, V150, and D2 of the rectum were also calculated on CTD30. Seed loss was determined 30 days after implantation. Urinary toxicity was measured with the International Prostate Symptom Score (IPSS) questionnaire.
Lower seed activity was associated with lower V150 and V200 (p = 0.01 and p ≤ 0.001, respectively) on CTD30. More patients had a V100 <90% and D90 <140 Gy in the lower activity cohort (p = 0.098 for D90 and p = 0.029 for V100) on CTD30. There was no difference between cohorts in dose to the rectum (p = 0.325-0.516) or difference in patients' IPSS score from baseline (p = 0.0.117-0.618), although there was a trend toward more urinary toxicity at 4 and 8 months for high activity seeds. Seed loss as a percentage of implanted seeds was not different (p = 0.324).
Higher seed activity (I(125) ≥ 0.6 mCi) results in at least equal V100 and D90 on CTD30. However, dose inhomogeneity and a trend toward more urinary toxicity at 4 and 8 months after treatment may lead to a higher long-term urinary complications.
International journal of radiation oncology, biology, physics 02/2010; 78(3):781-6. · 4.59 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: The low dose rate brachytherapy procedure would benefit from an intraoperative postimplant dosimetry verification technique to identify possible suboptimal dose coverage and suggest a potential reimplantation. The main objective of this project is to develop an efficient, operator-free, intraoperative seed detection technique using the imaging modalities available in a low dose rate brachytherapy treatment room.
This intraoperative detection allows a complete dosimetry calculation that can be performed right after an I-125 prostate seed implantation, while the patient is still under anesthesia. To accomplish this, a digital tomosynthesis-based algorithm was developed. This automatic filtered reconstruction of the 3D volume requires seven projections acquired over a total angle of 60 degrees with an isocentric imaging system.
A phantom study was performed to validate the technique that was used in a retrospective clinical study involving 23 patients. In the patient study, the automatic tomosynthesis-based reconstruction yielded seed detection rates of 96.7% and 2.6% false positives. The seed localization error obtained with a phantom study is 0.4 +/- 0.4 mm. The average time needed for reconstruction is below 1 min. The reconstruction algorithm also provides the seed orientation with an uncertainty of 10 degrees +/- 8 degrees. The seed detection algorithm presented here is reliable and was efficiently used in the clinic.
When combined with an appropriate coregistration technique to identify the organs in the seed coordinate system, this algorithm will offer new possibilities for a next generation of clinical brachytherapy systems.
Medical Physics 11/2009; 36(11):5235-44. · 2.83 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: Radiochromic film is a dosimeter of choice in applications requiring high spatial resolution, two dimensional measurements, or minimum perturbation of the beam fluence. Since the measurement uncertainty in Gafchromic film dosimetry is thought to be significant compared to that of ionization chambers, a rigorous method to evaluate measurement uncertainties is desired. This article provides a method that takes into account the correlation between fit parameters as well as single dose values in order to obtain accurate uncertainties in absolute and relative measurements. A complete portrait of all sources of uncertainty in Gafchromic film dosimetry is given. The parametrization of variance as a function of the number of averaged pixels is obtained in order to accurately predict the uncertainty as a function of the size of the region of interest. The choice of functional form for the sensitometric curve is based on four criteria and a convergence of global net optical density uncertainty to 0.0013 is demonstrated. A minimum number of 12 points is recommended to characterize the sensitometric curve to a sufficient precision on the uncertainty estimation. Uncertainty levels of 0.9% on absolute dose measurements and 0.45% on relative measurements are achieved using a 12-point calibration curve with 220 cGy and repeating measurements five times. Uncertainties of 0.8% and 0.4% are achievable when using 35 points during film characterization. Ignoring covariance terms is shown to lead to errors in the estimation of uncertainty.
Medical Physics 07/2009; 36(6):1931-46. · 2.83 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: To assess the impact of fiducial markers and daily kilovoltage imaging (FM-kV) on dose-volume histogram (DVH) parameters and normal tissue complication probabilities (NTCPs) for the rectum and bladder during prostate cancer radiotherapy.
Two different setup scenarios were compared for 20 patients treated with three-dimensional conformal radiotherapy (3D-CRT) for localized prostate cancer to a total dose of 76 Gy: a traditional setup with planning target volume (PTV) margins associated with skin mark alignment vs. another setup using FM-kV. Various DVH parameters were compared, including Radiation Therapy Oncology Group (RTOG) dose-volume constraints for the rectum and bladder. Analysis of NTCPs was also performed according to the Lyman model.
With the traditional setup, 85% of patients had rectal V70(Gy) >25% compared with 45% with FM-kV. Moreover, 30% of patients with traditional setup vs. 5% with FM-kV did not fulfill at least 3 RTOG constraint parameters for the rectum. Mean rectal and bladder dose were 4.7 Gy and 6.7 Gy less, respectively, with FM-kV. The NTCP for the rectum was 11.5% with the traditional setup and 9% with FM-kV. This indicates that with FM-kV, the prescription dose could be increased by 2.1 Gy while keeping the same level of late rectal toxicity as with the traditional setup.
Use of FM-kV is an efficient way of lowering the proportion of patients not fulfilling RTOG rectal and bladder dose-volume constraints. The results of the NTCP analysis suggest that the PTV margin reduction allowed by FM-kV should decrease the rate of late rectal toxicities or may allow moderate dose escalation.
International journal of radiation oncology, biology, physics 01/2009; 74(4):1128-33. · 4.59 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: To identify dosimetric parameters predictive of a good prostate seed I(125) quality implant. We analyzed preimplant and postimplant realtime dosimetry in patients treated with intraoperative (IO) inverse planning.
We analyzed 127 consecutively treated patients with primarily low-risk prostate carcinoma who underwent prostate permanent seed I(125) brachytherapy using an IO planning approach. The implant was done using the three-dimensional transrectal ultrasound (PRE-TRUS)-guided IO interactive inverse preplanning system. The TRUS was repeated in the operating room after the implant procedure was complete (POST-TRUS). The prostate was recontoured and postimplant dosimetry was calculated. Each patient underwent computed tomography scan on Day 28 (CT-D28) to evaluate implant quality. Area under the receiver operating characteristic curves (AUROC) was evaluated for models predictive of a V100 of > or =90% and a D90 of > or =140 Gy on the basis of CT-D28 values.
On CT-D28, 72.4% of patients had a V100 of > or =90% and 74.8% had a D90 of > or =140 Gy. AUROC for a V100 of > or =90% was 0.665 (p = 0.004) on PRE-TRUS and 0.619 (p = 0.039) on POST-TRUS. AUROC for D90 of > or =140 Gy was 0.602 (p = 0.086) on PRE-TRUS and 0.614 (p = 0.054) on POST-TRUS. Using PRE-TRUS V100 cutoff of >97% gives sensitivity of 88% and a false-positive rate of 63%. A POST-TRUS D90 cutoff of >170 Gy resulted in a sensitivity of 62% and a false-positive rate of 34%.
Because of unacceptably high false-positive rates, IO preimplant and postimplant TRUS-based dosimetry are not accurate tools to predict for postimplant computed tomography-based dosimetry.
International Journal of Radiation OncologyBiologyPhysics 11/2008; 72(2):605-9. · 4.11 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: Standard algorithms for postimplant analysis of transperineal interstitial permanent prostate brachytherapy (TIPPB) are based on AAPM Task Group 43 formalism (TG-43), which makes use of a world entirely made of water. This entails an assignment of the prostate, surrounding organs at risk, as well as all brachytherapy seeds present in a permanent prostate implant to water. Brachytherapy seeds are generally made from high atomic number materials. Because of the simultaneous presence of many brachytherapy seeds in a TIPPB, there is a shielding effect causing an attenuation of energy of the emitted photons generally called the "interseed attenuation" (ISA). This study investigates the impact of seed designs and compositions on the interseed attenuation. For this purpose, six brachytherapy seeds covering a wide variety of seed design and composition were modeled with the GEANT4 Monte Carlo (MC) toolkit. MC has allowed calculation of the contribution of each major component (encapsulation and internal components) of a given seed model to ISA separately. The impact of ISA on real clinical implant configurations was also explored. Two clinical postimplant geometries with different brachytherapy seeds were studied with MC simulations. The change in the clinical parameter D90 was observed. This study shows that Nucletron SelectSeed (similar to the Oncura model 6711), ProstaSeed, and Best Medical model 2335 are the most attenuating designs with 4.8%, 3.9%, and 4.6% of D90 reduction, respectively. The least attenuating seed is a 103Pd seed encapsulated in a polymer shell, the IBt OptiSeed with 1.5%. Finally, based on this systematic study, a new seed design is proposed that is predicted to be the most waterlike brachytherapy seed and thus TG-43 compatible.
Medical Physics 09/2008; 35(8):3671-81. · 2.83 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: Monte Carlo (MC) dose calculations are performed on patient geometries derived from computed tomography (CT) images. For most available MC codes, the Hounsfield units (HU) in each voxel of a CT image have to be converted into mass density (rho) and material type. This is typically done with a (HU; rho) calibration curve which may lead to mis-assignment of media. In this work, an improved material segmentation using dual-energy CT-based material extraction is presented. For this purpose, the differences in extracted effective atomic numbers Z and the relative electron densities rho(e) of each voxel are used. Dual-energy CT material extraction based on parametrization of the linear attenuation coefficient for 17 tissue-equivalent inserts inside a solid water phantom was done. Scans of the phantom were acquired at 100 kVp and 140 kVp from which Z and rho(e) values of each insert were derived. The mean errors on Z and rho(e) extraction were 2.8% and 1.8%, respectively. Phantom dose calculations were performed for 250 kVp and 18 MV photon beams and an 18 MeV electron beam in the EGSnrc/DOSXYZnrc code. Two material assignments were used: the conventional (HU; rho) and the novel (HU; rho, Z) dual-energy CT tissue segmentation. The dose calculation errors using the conventional tissue segmentation were as high as 17% in a mis-assigned soft bone tissue-equivalent material for the 250 kVp photon beam. Similarly, the errors for the 18 MeV electron beam and the 18 MV photon beam were up to 6% and 3% in some mis-assigned media. The assignment of all tissue-equivalent inserts was accurate using the novel dual-energy CT material assignment. As a result, the dose calculation errors were below 1% in all beam arrangements. Comparable improvement in dose calculation accuracy is expected for human tissues. The dual-energy tissue segmentation offers a significantly higher accuracy compared to the conventional single-energy segmentation.
Physics in Medicine and Biology 06/2008; 53(9):2439-56. · 2.83 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: Tissue segmentation is an important step in Monte Carlo (MC) dose calculation and is often done uncritically. A new approach to tissue segmentation using dual-energy CT images is studied in this work.
A simple MC model of a CT scanner was built and CT images of phantoms with ten tissue-equivalent cylinders were simulated using soft and hard X-ray spectra. The Z and rho(e) of the cylinders were extracted using a formalism based on a parameterization of the linear attenuation coefficient.
It was shown that in order to extract Z and rho(e) with a reasonable accuracy, hard X-ray beams have to be used for scanning. When an additional filtration of 9 mm of aluminium in the CT X-ray beam is used, beam hardening in high density materials is suppressed and the mean errors of the extraction of Z and rho(e) for 10 tissue-equivalent materials in a small tissue-equivalent phantom are 3.7% and 3.1%, respectively.
MC simulations were used to show that the extraction of Z and rho(e) for a number of tissue-equivalent materials using dual-energy CT images is possible which improves tissue segmentation for Monte Carlo dose calculations, as demonstrated with a 250 kVp photon beam dose calculation.
Radiotherapy and Oncology 02/2008; 86(1):93-8. · 5.58 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: To use the Monte Carlo (MC) method as a dose calculation engine for postimplant dosimetry. To compare the results with clinically approved data for a sample of 28 patients. Two effects not taken into account by the clinical calculation, interseed attenuation and tissue composition, are being specifically investigated.
An automated MC program was developed. The dose distributions were calculated for the target volume and organs at risk (OAR) for 28 patients. Additional MC techniques were developed to focus specifically on the interseed attenuation and tissue effects.
For the clinical target volume (CTV) D(90) parameter, the mean difference between the clinical technique and the complete MC method is 10.7 Gy, with cases reaching up to 17 Gy. For all cases, the clinical technique overestimates the deposited dose in the CTV. This overestimation is mainly from a combination of two effects: the interseed attenuation (average, 6.8 Gy) and tissue composition (average, 4.1 Gy). The deposited dose in the OARs is also overestimated in the clinical calculation.
The clinical technique systematically overestimates the deposited dose in the prostate and in the OARs. To reduce this systematic inaccuracy, the MC method should be considered in establishing a new standard for clinical postimplant dosimetry and dose-outcome studies in a near future.
International Journal of Radiation OncologyBiologyPhysics 08/2007; 68(4):1190-8. · 4.11 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: The purpose is to evaluate the impact of interseed attenuation and prostate composition for prostate treatment plans with 125I permanent seed implants using the Monte Carlo (MC) method. The effect of seed density (number of seeds per prostate unit volume) is specifically investigated. The study focuses on treatment plans that were generated for clinical cases. For each plan, four different dose calculation techniques are compared: TG-43 based calculation, superposition MC, full MC with water prostate, and full MC with realistic prostate tissue. The prostate tissue description is from the ICRP report 23 (W. S. Snyer, M. J. Cook, E. S. Nasset, L. R. Karkhausen, G. P. Howells, and I. H. Tipton, "Report of the task group on reference man," Technical Report 23, International Commission on Radiological Protection, 1974). According to the comparisons, the seed density has an influence on interseed attenuation. A plan with a typical low seed density (42 0.6 mCi seeds in a 26 cm3 prostate) suffers a 1.2% drop in the CTV D90 value due to interseed attenuation. A drop of 3.0% is calculated for a higher seed density (75 0.3 mCi seeds, same prostate). The influence of the prostate composition is similar for all seed densities and prostate sizes. The difference between MC simulations in water and MC simulations in prostate tissue is between 4.4% and 4.8% for the D90 parameter. Overall, the effect on D90 is ranging from 5.8% to 12.8% when comparing clinically approved TG-43 and MC simulations in prostate tissue. The impact varies from one patient to the other and depends on the prostate size and the number of seeds. This effect can reach a significant level when reporting correlations between clinical effect and deposited dose.
Medical Physics 04/2006; 33(3):595-604. · 2.83 Impact Factor
