Calculate Organ Doses Research Articles

DEVELOPMENT OF RAPID METHODS FOR ASSESSING DOSES FROM INTERNALLY DEPOSITED RADIONUCLIDES

During the initial phases of the National Institute for Occupational Safety and Health Radiation Dose Reconstruction Program, all calculations of organ doses due to internally deposited radionuclides were performed using the Integrated Modules for Bioassay Analysis program. However, limitations associated with this program, including the need to calculate separate internal dose assessments for each radionuclide, created inefficiencies in the processing of claims. As a result, the National Institute for Occupational Safety and Health developed and introduced a suite of tools to expedite the process. The first of these was the Chronic Annual Dose Workbook program. This innovative tool permits a dose reconstructor to calculate, in a single step, an organ dose that involves up to 255 separate intakes in any combination of radionuclides, intake modes, and absorption types. In addition, the program enables dose reconstructors to determine the specific radionuclide characteristics that will deliver the highest organ dose for a specific intake. Furthermore, the results are displayed in a format that is compatible with the Interactive RadioEpidemiological Program, which is used by the U.S. Department of Labor in establishing the probability of causation. The value of the probability of causation, in combination with other information, subsequently enables the U.S. Department of Labor to render a decision on compensability. These developments have played a major role in enabling the dose reconstruction teams to meet the claim processing goals with increased efficiency and accuracy.

Organ dose calculations by Monte Carlo modeling of the updated VCH adult male phantom against idealized external proton exposure

The voxel-based visible Chinese human (VCH) adult male phantom has offered a high-quality test bed for realistic Monte Carlo modeling in radiological dosimetry simulations. The phantom has been updated in recent effort by adding newly segmented organs, revising walled and smaller structures as well as recalibrating skeletal marrow distributions. The organ absorbed dose against external proton exposure was calculated at a voxel resolution of 2 × 2 × 2 mm3 using the MCNPX code for incident energies from 20 MeV to 10 GeV and for six idealized irradiation geometries: anterior–posterior (AP), posterior–anterior (PA), left-lateral (LLAT), right-lateral (RLAT), rotational (ROT) and isotropic (ISO), respectively. The effective dose on the VCH phantom was derived in compliance with the evaluation scheme for the reference male proposed in the 2007 recommendations of the International Commission on Radiological Protection (ICRP). Algorithm transitions from the revised radiation and tissue weighting factors are accountable for approximately 90% and 10% of effective dose discrepancies in proton dosimetry, respectively. Results are tabulated in terms of fluence-to-dose conversion coefficients for practical use and are compared with data from other models available in the literature. Anatomical variations between various computational phantoms lead to dose discrepancies ranging from a negligible level to 100% or more at proton energies below 200 MeV, corresponding to the spatial geometric locations of individual organs within the body. Doses show better agreement at higher energies and the deviations are mostly within 20%, to which the organ volume and mass differences should be of primary responsibility. The impact of body size on dose distributions was assessed by dosimetry of a scaled-up VCH phantom that was resized in accordance with the height and total mass of the ICRP reference man. The organ dose decreases with the directionally uniform enlargement of voxels. Potential pathways to improve the VCH phantom have also been briefly addressed. This work pertains to VCH-based systematic multi-particle dose investigations and will contribute to comparative dosimetry studies of ICRP standardized voxel phantoms in the near future.

A study on the magnetic resonance imaging (MRI)-based radiation treatment planning of intracranial lesions

The aim of this study is to develop a magnetic resonance imaging (MRI)-based treatment planning procedure for intracranial lesions. The method relies on (a) distortion correction of raw magnetic resonance (MR) images by using an adaptive thresholding and iterative technique, (b) autosegmentation of head structures relevant to dosimetric calculations (scalp, bone and brain) using an atlas-based software and (c) conversion of MR images into computed tomography (CT)-like images by assigning bulk CT values to organ contours and dose calculations performed in Eclipse (Philips Medical Systems). Standard CT + MRI-based and MRI-only plans were compared by means of isodose distributions, dose volume histograms and several dosimetric parameters. The plans were also ranked by using a tumor control probability (TCP)-based technique for heterogeneous irradiation, which is independent of radiobiological parameters. For our 3 T Intera MRI scanner (Philips Medical Systems), we determined that the total maximum image distortion corresponding to a typical brain study was about 4 mm. The CT + MRI and MRI-only plans were found to be in good agreement for all patients investigated. Following our clinical criteria, the TCP-based ranking tool shows no significant difference between the two types of plans. This indicates that the proposed MRI-based treatment planning procedure is suitable for the radiotherapy of intracranial lesions.

SU‐GG‐I‐63: Calculations of Organ Doses From Cone‐Beam CT IGRT Procedures Using a Suite of Patient Phantoms

Purpose: Widespread uses of Computed Tomography (CT) and image‐guided radiation treatment (IGRT) can significantly increase the cumulative exposure to health tissues. AAPM TG‐75 was formed to compile data on imaging exposures and to identify ways to optimize the imaging procedure. To address the lack of organ doses and effective doses assessment in the IGRT imaging procedure, this paper describes a study to use a Monte Carlo based method to simulate kV and MV cone beam CT (CBCT) scanners and then calculate organ doses and effective doses. Method and Materials: This paper utilizes a whole‐body adult male phantom called the VIP‐Man, and Adult Male and Female that are adjusted to match with 50th percentile ICRP average population. The MCNPX code is used to simulate CT scanners and finally calculate doses. Several IGRT procedures are considered to demonstrate how this method can assess and compare various protocols. Results: For the VIP‐Man phantom, the effective dose for kV CBCT (assuming total exposure of 1350 mAs) is about 9.5 mSv for the two anatomical sites while the effective dose for MV CBCT (assuming a total of 6 MU) ranges from 5.10 mSv for the H&N case to 8.39 mSv for the prostate case. For different procedures, the effective dose data can be adjusted accordingly. Conclusion: With realistic patient phantoms and imaging scenario, organ dosees can be obtained by Monte Carlo simulation, and consequently effective doses have been calculated. The results show that showed that the effective dose is a useful quantity for the comparison of the performance and patient safety between kV CBCT and MV CBCT. These information can aid the assessment of radiation risk to the patient from imaging procedure. This study demonstrates the feasibility of determining detailed organ doses for various IGRT related cone‐beam CT procedures.

WE-D-332-04: Calculation of Patient-Specific Organ Doses From CT Exams

Purpose: To develop a Monte Carlo based package for calculating the patient specific organ doses from CT exams. Method and Materials: Clinical CT systems were simulated using the recently released Geant4 9.1 based on specifications provided by CT vendors. Continuous movements of CT gantry rotation and patient table translation were implemented. A text file was used to configure the scanner geometry and scanning protocols. The X-ray spectra were generated using the TASMIP program and were verified with the scanner HVL specifications. The simulation was at first validated by comparing the calculated CTDI100 with the measured value. The CT images of a CIRS anthropomorphic phantom were then used as a voxelized phantom and the CT ♯ of each voxel was converted to a tissue composition based on the published data. Absorbed energy in each voxel was recorded and 3-D dose map was obtained. Organs were segmented manually and the dose for each organ was calculated. Results: The difference between the simulated CTDI100 and the measured value was less than 10%. A CT scan of a CIRS anthropomorphic phantom was simulated and doses to each organ were calculated. Conclusion: It is feasible to calculate organ doses for individual patient based on clinically acquired CT images and scanning parameters. Automated segmentation tool is needed to make the organ dose calculation to be practical.

WE-B-351-02: Patient Dose Estimates From CT Examinations

Radiation dose from computed tomography has been an important issue for medical physicists for some time. This issue has increased in significance in the past several years due to advances in multidetector CT, which have resulted in increased utilization in areas such as pediatric CT, cardiac CT and even screening applications. Some of the key issues currently facing the medical physics community are assessing dose to patients from these various exams. One of the key building blocks to these assessments is the estimation of organ dose. The purpose of this presentation is to describe an approach to estimating organ doses to patients using Monte Carlo-based simulation methods. In this approach, both the scanner and patient are modeled in some detail and a CT exam may be simulated. The detailed modeled of the CT scanner is created by including information such as the source spectra and filtration, its geometry, beam collimation and the path that the source travels around the patient (such as the path during a helical CT exam). The development, testing and validation of these models will be discussed. Patient models generally fall into two categories. The first consists of geometric descriptions of organs (based on cones, cylinders, etc.) such as the MIRD phantom. The second consists of voxelized descriptions of patient anatomy that are created based on actual patient scans. In these, radiosensitive organs are identified in the image data to create a voxelized model of the patient geometry. For both types of models, there are challenges to create models representing patients of different sizes, ages and genders. Once both the scanner and patient are modeled, then different scan protocols can be simulated using a Monte Carlo based software package (such as MCNP or EGS). This involves selecting a scanner model, a patient model and then selecting a set of technical parameters, such as one would do for an actual scan — including body region being examined, etc. The Monte Carlo software then simulates the specified scan and tallies absorbed dose in each voxel or geometric unit of the patient model, which then allows the calculation of either mean organ dose or the distribution of dose within an organ. In this presentation, the results of this approach in several applications will be described, including: (a) dose to the fetus in pregnant women of early, middle and late gestational ages, (b) comparing dose to glandular breast tissue from thoracic CT scans both with and without tube current modulation. Educational Objectives: 1. Understand the Monte Carlo simulation based approach to estimating radiation dose to radiosensitive organs from CT scans. 2. Understand the current limitations of the Monte Carlo based approach. 3. Describe the results of some current applications of this approach to estimating fetal dose as well as breast dose reduction from tube current modulation.

SU‐GG‐T‐347: Methods to Calculate Organ Doses Due to Scattered Radiation From External Beam Radiation Treatment Using Various Patient Phantoms and Monte Carlo Simulations

Purpose: It has been long known that patients treated with radiation carry a risk of developing second cancers in their lifetimes as a result of their treatments. In addition to hosting a symposium on “Secondary Cancer Risks for Emerging Radiation Treatments,” the AAPM has recently formed Task Group 158 that will provide recommendations on the dosimetry protocols for quantifying scattered radiation dose outside the treatment volume. Measurements are useful in providing a relative comparison of different treatment modalities or beam characteristics. On the other hand, more precise organ dose equivalent values need to be calculated from Monte Carlo simulations and anatomically realistic computational phantoms. This study outlines a computational platform that integrates anatomically‐refined virtual patient models with a detailed model of a medical accelerator head to calculate scattered doses to patients treated with radiation. Method and Materials: A detailed model of a 6‐MV Varian Clinac 2100C was developed in MCNPX. Several virtual patient models have been integrated with the accelerator head model in MCNPX. Various conventional and IMRT treatment plans have been considered. Results: Average organ doses are significantly different than point‐wise measurement values. The differences can be attributed to various sources of scattered radiation that deposits dose in the organ. Large dose gradients will be present throughout organs at short and intermediate distances from the tumor volume, while organs at longer distances will have a more uniform dose distribution. Conclusion: A computational platform to calculate organ dose from scattered radiation during radiation treatments has been demonstrated. Results indicate that average organ/tissue doses are significantly different than point‐wise measurements. Organ doses calculated in this manner will likely provide more accurate organ dose information for epidemiologic studies and for assessing the risk during the treatment planning.

Dosimetric models used in the Alpha-Risk project to quantify exposure of uranium miners to radon gas and its progeny

The European project Alpha-Risk aims to quantify the cancer and non-cancer risks associated with multiple chronic radiation exposures by epidemiological studies, organ dose calculation and risk assessment. In the framework of this project, mathematical models have been applied to the organ dosimetry of uranium miners who are internally exposed to radon and its progeny as well as to long-lived radionuclides present in the uranium ore. This paper describes the methodology and the dosimetric models used to calculate the absorbed doses to specific organs arising from exposure to radon and its progeny in the uranium mines. The results of dose calculations are also presented.

Calculation of organ doses in x-ray examinations of premature babies

Lung disease represents one of the most life-threatening conditions in prematurely born children. In the evaluation of the neonatal chest, the primary and most important diagnostic study is the chest radiograph. Since prematurely born children are very sensitive to radiation, those radiographs may lead to a significant radiation detriment. Knowledge of the radiation dose is therefore necessary to justify the exposures. To calculate doses in the entire body and in specific organs, computational models of the human anatomy are needed. Using medical imaging techniques, voxel phantoms have been developed to achieve a representation as close as possible to the anatomical properties. In this study two voxel phantoms, representing prematurely born babies, were created from computed tomography- and magnetic resonance images: Phantom 1 (1910 g) and Phantom 2 (590 g). The two voxel phantoms were used in Monte Carlo calculations (MCNPX) to assess organ doses. The results were compared with the commercially available software package PCXMC in which the available mathematical phantoms can be downsized toward the prematurely born baby. The simple phantom-scaling method used in PCXMC seems to be sufficient to calculate doses for organs within the radiation field. However, one should be careful in specifying the irradiation geometry. Doses in organs that are wholly or partially outside the primary radiation field depend critically on the irradiation conditions and the phantom model.

Construction of a Voxel Phantom Based on CT Data for a Japanese Female Adult and Its Use for Calculation of Organ Doses from External Electrons

Construction of a Voxel Phantom Based on CT Data for a Japanese Female Adult and Its Use for Calculation of Organ Doses from External Electrons

Construction of a Voxel Phantom Based on CT Data for a Japanese Female Adult and Its Use for Calculation of Organ Doses from External Electrons

A voxel phantom for a Japanese female adult was developed from CT data. This new phantom, Onago, has a slightly larger size than the Asian Reference Man, Female (ARMF). The organ doses and effective doses for external electrons were calculated with Onago in the energy range of 1MeV through 10GeV, and compared with those of a Japanese male voxel phantom Otoko and with those of a MIRD phantom. The organ doses show some typical tendencies which are common to all phantoms, and the absolute values are also similar. However, in some cases the absolute values of organ doses differ greatly among the phantoms: the maximum difference observed between the voxel phantoms and the MIRD-type phantom was a factor of fifty for the individual organ dose and a factor of two for the effective dose. This suggests the necessity of sophisticated phantoms and detailed investigation using the phantoms to find out the realistic reference organ doses and the fluctuations due to different factors for external electron exposures.

Body size-dependent patient effective dose for diagnostic radiography

The health-related mean absorbed dose to organs and tissues and the effective dose were usually calculated from the entrance surface dose (ESD) using computational phantoms by the Monte Carlo method. Because these phantoms represented an average adult with simplified organ shape and position within the body, the calculated results were hypothetical doses to the phantom but not to the patient. In order to determine the patient doses, the body size of individual patients should be taken into consideration. In the present work, two approaches were adopted to estimate the radiological thickness of a patient which determined the overlying tissue shielding and photon attenuation in the calculation of organ and effective doses. The first approach involved the determination of this thickness from patient body weight and height. It was found that the body mass index (BMI) was approximately linearly proportional to the body thickness. Therefore, the BMI was used to determine the size-dependent organ and effective doses from patient ESDs measured by the thermoluminescent dosimeters. To verify the applicability of BMI, WinODS software was employed to compute the organ and effective doses using the size- and sex-adjustable anthropomorphic phantom. The second approach involved the estimate of body thickness from the transmission factor, defined as the ratio of exit surface dose to ESD. The transmission factor was measured using polyethylene phantoms of varying thicknesses. Also, a backscatter factor, defined as the ratio of ESDs with and without the presence of patient, was measured using polyethylene phantoms of varying thicknesses.

TU‐SAMS‐L100J‐02: Determination of Dose From CT Examinations: Estimating Organ Dose Using Monte Carlo‐Based Methods

Radiation dose from computed tomography has been an important issue for medical physicists for some time. This issue has increased in significance in the past several years due to advances in multidetector CT, which have resulted in increased utilization in areas such as pediatric CT, cardiac CT and even screening applications. Some of the key issues currently facing the medical physics community are assessing dose to patients from these various exams. One of the key building blocks to these assessments is the estimation of organ dose.The purpose of this presentation is to describe an approach to estimating organ doses to patients using Monte Carlo‐based simulation methods. In this approach, both the scanner and patient are modeled in some detail and a CT exam may be simulated.The detailed modeled of the CT scanner is created by including information such as the source spectra and filtration, its geometry, beam collimation and the path that the source travels around the patient (such as the path during a helical CT exam). The development, testing and validation of these models will be discussed.Patient models generally fall into two categories. The first consists of geometric descriptions of organs (based on cones, cylinders, etc.) such as the MIRD phantom. The second consists of voxelized descriptions of patient anatomy that are created based on actual patient scans. In these, radiosensitive organs are identified in the image data to create a voxelized model of the patient geometry. For both types of models, there are challenges to create models representing patients of different sizes, ages and genders.Once both the scanner and patient are modeled, then different scan protocols can be simulated using a Monte Carlo based software package (such as MCNP or EGS). This involves selecting a scanner model, a patient model and then selecting a set of technical parameters, such as one would do for an actual scan — including body region being examined, etc. The Monte Carlo software then simulates the specified scan and tallies absorbed dose in each voxel or geometric unit of the patient model, which then allows the calculation of either mean organ dose or the distribution of dose within an organ.In this presentation, the results of this approach in several applications will be described, including: (a) dose to the fetus in pregnant women of early, middle and late gestational ages, (b) comparing dose to glandular breast tissue from thoracic CT scans both with and without tube current modulation.Educational Objectives:1. Understand the Monte Carlo simulation based approach to estimating radiation dose to radiosensitive organs from CT scans.2. Understand the current limitations of the Monte Carlo based approach.3. Describe the results of some current applications of this approach to estimating fetal dose as well as breast dose reduction from tube current modulation.

WE‐C‐AUD‐06: Radiation Safety of Pregnant Patients During Radiation Treatment: A Detailed Modeling of the Accelerator, Patient Anatomy, and Non‐Target Doses

Purpose: The cancer incidence rate in pregnant women is increasing, due to both the new trend for delaying pregnancy into late reproductive ages and early detection of common cancers associated with pregnancy. When radiation treatment is chosen, the safety concern for the mother and fetus must be fully addressed. However, previous studies of fetal dose, such as those reported by AAPM TG‐36, had been limited to homogenous water phantoms and surface measurements that are not quantitative. This paper presents our effort to develop medical accelerators and ICRP standard reference pregnant patient models for Monte Carlo calculations of nontarget doses during radiation treatments. Method and Materials: The Monte Carlo program MCNPX was used to develop a complete model of a Varian Clinac 2100C. Peripheral dose profiles in a water phantom were investigated for the following square fields defined by the jaws: 5 cm × 5 cm, 10 cm × 10 cm, and 20 cm × 20 cm. Anatomically‐realistic pregnant patient and fetal models at different stages of gestation representing ICRP reference female patients were developed using latest anatomical modeling technologies. Results: The peripheral dose characteristics as a function of distance from the field edge for 5 cm × 5 cm, 10 cm × 10 cm, and 20 cm × 20 cm field sizes at several depths are consistent with previous studies. The change in peripheral dose as a function of depth is small for each field size. The peripheral dose increases as field size increases. Conclusion: Calculations of organ doses to a pregnant patient and fetus have been demonstrated. Such data will help better assess the risks to the mother and fetus from radiation treatment procedures. The tools we have developed can be adopted for routine use to optimize the treatment planning for pregnant patients.

Specific absorbed fraction for korean adult voxel phantom from internal photon source

Absorbed fraction (AF) and specific absorbed fraction (SAF) are crucial values for the calculation of radionuclide S-values and consequently for internal dose estimates. The formalism of the Medical Internal Radiation Dose (MIRD) committee of the Society of Nuclear Medicine (SNM) has been utilised as a standard in the calculation of individual organ doses for biologically distributed radionuclides and for different types of radiation. Although those quantities are highly sensitive to individual anatomical difference, the SAF dataset calculated by Caucasian-based stylised phantoms have been applied to Korean population until now. This study was intended to calculate the SAFs by using realistic Korean voxel phantom and Monte Carlo transport technique for the first time and compare the results with those of the existing Caucasian-based data and the Korean stylised phantom published recently. The up-to-date realistic Korean voxel phantom, KTMAN-2, which was developed from computed tomography (CT) images of an average Korean adult male, was employed for Monte Carlo calculation using EGSnrc user-code, developed for the purpose of this study. The SAFs for 32 target organs and tissues from the photon source, uniformly deposited in a total of 37 source organs and tissues, were calculated from KTMAN-2. The results were compared with those for an adult phantom of Oak Ridge National Laboratory (ORNL) and Korean adult stylised phantom. Two major reasons of discrepancy were analysed: (1) racial difference between the Korean and the Caucasian and (2) anatomical difference between stylised and voxel phantoms. When the source organ was identical to the target organ, difference in SAF caused by the difference in target-organ mass between the Korean and the Caucasian phantoms was mainly observed. When the source and target organs were not identical, significant difference in SAF was observed which was mainly attributed to the difference in inter-organ distance and organ shape between voxel and stylised phantoms.

Quantification of Cancer and Noncancer Risks Associated With Multiple Chronic Radiation Exposures: Epidemiologic Studies, Organ Dose Calculation, and Risk Assessment (Alpha-Risk)

Quantification of Cancer and Noncancer Risks Associated With Multiple Chronic Radiation Exposures: Epidemiologic Studies, Organ Dose Calculation, and Risk Assessment (Alpha-Risk)

Radiation Exposure to Patient and Staff in Hepatic Chemoembolization: Risk Estimation of Cancer and Deterministic Effects

The purpose of the study was to determine the risks of radiation-induced cancer and deterministic effects for the patient and staff in transarterial chemoembolization (TACE) of hepatocellular carcinoma (HCC). Sixty-five patients with HCC underwent the first cycle of TACE. Thermoluminescence dosemeters and conversion factors were used to measure surface doses and to calculate organ doses and effective dose. For the patient, the risk of fatal cancer and severe genetic defect was in the magnitude of 10(-4) and 10(-5), respectively. Five patients showed surface doses over the first lumbar vertebra exceeding 2000 mSv and 45 patients showed doses over the spine or the liver region above 500 mSv. The risk of fatal cancer and severe genetic defect for the radiologist and assistant was in the magnitude of 10(-7) to 10(-8). They could exceed the threshold for lens opacity in the case of more than 490 and 1613 TACE yearly for a period of many years, respectively. Radiation dose could lead to local transient erythema and/or local depression of hematopoiesis in many patients after TACE. For the radiologist and assistant, risk of fatal cancer and genetic defect and lens opacity might arise when they perform interventions such as TACE intensively.

A tomographic physical phantom of the newborn child with real-time dosimetry. II. Scaling factors for calculation of mean organ dose in pediatric radiography

Following the recent completion of a tomographic physical newborn dosimetry phantom with incorporated metal-oxide-semiconductor field effect transistor (MOSFET) dosimetry system, it was necessary to derive scaling factors in order to calculate organ doses in the physical phantom given point dose measurements via the MOSFET dosimeters (preceding article in this issue). In this study, we present the initial development of scaling factors using projection radiograph data. These point-to-organ dose scaling factors (SF(POD)) were calculated using a computational phantom created from the same data set as the physical phantom, but which also includes numerous segmented internal organs and tissues. The creation of these scaling factors is discussed, as well as the errors associated when using only point dose measurements to calculate mean organ doses and effective doses in physical phantoms. Scaling factors for various organs ranged from as low as 0.70 to as high as 1.71. Also, the ability to incorporate improvements in the computational phantom into the physical phantom using scaling factors is discussed. An comprehensive set of SF(POD) values is presented in this article for application in pediatric radiography of newborn patients.

EFFECTIVE DOSE EQUIVALENT FOR POINT GAMMA SOURCES LOCATED 10 CM NEAR THE BODY

The key component in the so-called EPRI effective dose equivalent (EDE) methodology is an algorithm that utilizes two dosimeters (instead of multiple dosimeters) to predict the EDE for external photon exposures. The exposure scenarios that were previously studied in deriving the algorithm include parallel photon beams and point sources 33 cm from the body surface. The motivation for this study was the need to investigate source locations within 33 cm from the body so the method is more widely applicable. The ORNL stylized mathematical human phantoms and the MCNP code were used to calculate organ doses in this study. This paper presents the EDE data for point gamma sources at 0.3, 1.0, and 1.5 MeV, respectively, which are located at 10 cm from the surface of the body. The results and analyses show that the locations ranging from the overhead to the foot have resulted in conservative ratios except for two general regions near the front upper thigh and directly overhead. If all locations considered in this study were averaged for each photon energy, the overall ratio is on the conservative side. These data suggest that the EPRI EDE methodology is still valid for sources located 10 cm from the body, although the chance for resulting in a non-conservative estimate of the EDE has increased in comparison with the sources located at 30 cm from the body. Finally, this paper provides recommendations on how to apply the EPRI EDE methodology.

WE‐E‐224C‐04: A New CT Reconstruction Technique for Removal of Streak Artifacts Due to Metallic Dental Fillings and Implants for the Treatment of Head and Neck Cancer with Intensity Modulated Radiation Therapy

Purpose: To remove or reduce the streak artifacts induced by metallic dental fillings and implants in CT images for head and neck patients treated with intensity modulated radiation therapy (IMRT). Materials and Method: A cylindrical phantom was constructed using tissue‐equivalent plastic to simulate a human head. The phantom was filled with distilled water. A piece of 1 cm‐wide half elliptic tissue‐equivalent bolus was attached to a thin circular plastic plate inside the phantom to mimic the gingiva. Two human second molars with metallic fillings were implanted into the bolus on each side. The phantom was scanned on a GE LightSpeed CT scanner with a slice thickness of 2.5 mm. The acquired CT images were first decomposed into spatio‐frequency components using a wavelet transform. The decomposition tree allowed us to examine and characterize different localized frequency information related to undesired artifacts. Once the spatio‐frequency signatures of metal objects were modeled, the corresponding coefficients in the wavelet domain were thresholded using a dynamic thresholding scheme. The thresholded images gave metal objets, which were then superimposed on a neighbor slice. New data were generated by re‐projecting the original slice and superimposed slice. The final image was reconstructed with no or reduced image artifacts due to dental fillings and implants. Results: Comparison of original images to reconstructed images indicates that the streak artifacts were either completely removed or significantly reduced. The structural details near the metal objects were clearly identifiable. The image quality was sufficient for target delineation in IMRT treatment planning. Conclusions: The proposed technique can greatly reduce the streak artifacts induced by metallic dental fillings and implants without sacrifice of spatial resolution. The technique has the potential to significantly improve the accuracy of target and critical organ delineation and dose calculation in the head and neck IMRT treatment planning.