Monte Carlo simulation of the Leksell Gamma Knife: I. Source modelling and calculations in homogeneous media

The Monte Carlo code PENELOPE has been used to simulate photon flux from the Leksell Gamma Knife, a precision method for treating intracranial lesions. Radiation from a single 6OCo assembly traversing the collimator system was simulated, and phase space distributions at the output surface of the helmet for photons and electrons were calculated. The characteristics describing the emitted final beam were used to build a two-stage Monte Carlo simulation of irradiation of a target. A dose field inside a standard spherical polystyrene phantom, usually used for Gamma Knife dosimetry, has been computed and compared with experimental results, with calculations performed by other authors with the use of the EGS4 Monte Carlo code, and data provided by the treatment planning system Gamma Plan. Good agreement was found between these data and results of simulations in homogeneous media. Owing to this established accuracy, PENELOPE is suitable for simulating problems relevant to stereotactic radiosurgery.     

Calibration of the Gamma Knife using a new phantom following the AAPM TG51 and TG21 protocols

PURPOSE: To compare calibration of the Leksell Gamma Knife according to the American Association of Physicists in Medicine Task Groups 21 and 51 protocols. A new phantom was fabricated for this purpose. Its design, physical properties, and composition are described. MATERIALS AND METHODS: The Gamma Knife TG-51 calibration phantom is designed to be filled with water and support an ionization chamber positioned at its center. The phantom is thimble-shaped, with a 2 mm plastic wall to contain water. The phantom and chamber assembly was mounted in a Leksell stereotactic frame. The location of the chamber's sensitive volume was determined using computed tomography. The chamber-phantom assembly was attached to the 18 mm helmet in the Gamma Knife by the stereotactic frame. The phantom's geometry allowed radiation beams from each of the 201 Gamma Knife cobalt-60 sources to converge after an 8 cm path to the ionization chamber's sensitive volume. This is similar to the arrangement by which one calibrates the Gamma Knife using the manufacturer-supplied polystyrene phantom. RESULTS: The phantom was attached to the Gamma Knife so that the ionization chamber was reproducibly positioned at the convergence of the radiation beams. Because of the phantom's design, the phantom could be affixed to either trunnions or the automatic patient positioning system, once mounted in the Leksell stereotectic frame. Comparisons using different phantoms and protocols resulted in the following calibration ratios for TG-21 in the polystyrene sphere phantom, TG-21 in the water phantom, and TG-51 in the water phantom, respectively: 1.000, 1.008, 0.986, when corrected for transmission through the plastic water reservoir wall and using the same ionization chamber. Transmission measurements using a 1 cm thickness of the same material in the Co-60 beam determined that the phantom's 2 mm plastic wall resulted in a reduction in the measured the output by 0.5%. CONCLUSIONS: Calibration of the Gamma Knife can be performed in liquid water using the AAPM TG-51 protocol and this new phantom, thereby eliminating uncertainties with respect to the composition of the manufacturer's phantom. Perturbation of calibration measurements by nonwater materials was characterized and could be corrected. Calibration values for the Gamma Knife that were obtained using the three methods for our phantoms agree to within 1.4%. TG21 and TG51 calibration of the Gamma Knife using the water phantom agreed to within 2.2%.     

MCNP-based computational model for the Leksell gamma knife

We have focused on the usage of MCNP code for calculation of Gamma Knife radiation field parameters with a homogenous polystyrene phantom. We have investigated several parameters of the Leksell Gamma Knife radiation field and compared the results with other studies based on EGS4 and PENELOPE code as well as the Leksell Gamma Knife treatment planning system Leksell GammaPlan (LGP). The current model describes all 201 radiation beams together and simulates all the sources in the same time. Within each beam, it considers the technical construction of the source, the source holder, collimator system, the spherical phantom, and surrounding material. We have calculated output factors for various sizes of scoring volumes, relative dose distributions along basic planes including linear dose profiles, integral doses in various volumes, and differential dose volume histograms. All the parameters have been calculated for each collimator size and for the isocentric configuration of the phantom. We have found the calculated output factors to be in agreement with other authors' works except the case of 4 mm collimator size, where averaging over the scoring volume and statistical uncertainties strongly influences the calculated results. In general, all the results are dependent on the choice of the scoring volume. The calculated linear dose profiles and relative dose distributions also match independent studies and the Leksell GammaPlan, but care must be taken about the fluctuations within the plateau, which can influence the normalization, and accuracy in determining the isocenter position, which is important for comparing different dose profiles. The calculated differential dose volume histograms and integral doses have been compared with data provided by the Leksell GammaPlan. The dose volume histograms are in good agreement as well as integral doses calculated in small calculation matrix volumes. However, deviations in integral doses up to 50% can be observed for large volumes such as for the total skull volume. The differences observed in treatment of scattered radiation between the MC method and the LGP may be important in this case. We have also studied the influence of differential direction sampling of primary photons and have found that, due to the anisotropic sampling, doses around the isocenter deviate from each other by up to 6%. With caution about the details of the calculation settings, it is possible to employ the MCNP Monte Carlo code for independent verification of the Leksell Gamma Knife radiation field properties.     

The use of a Leksell-BRW adapter for linac radiosurgery as an adjunct to Gamma Knife treatment

We have investigated the use of an adapter that permits the use of a Leksell coordinate frame with a linear accelerator stereotactic radiosurgery system based on the Brown-Robert-Wells (BRW) design. This device is useful when lesions that are planned for treatment on a Leksell Gamma Knife system are found to be inaccessible to the Gamma Knife. We have found that with this device objects within a head phantom can be targeted by the linear accelerator within an accuracy of approximately 1 mm.     

Accuracy and stability of positioning in radiosurgery: long-term results of the Gamma Knife system

The primary aim of this investigation was to determine the long term overall accuracy of an irradiation position of Gamma Knife systems. The mechanical accuracy of the system as well as the overall accuracy of an irradiation position was examined by irradiating radiosensitive films. To measure the mechanical accuracy, the GafChromic film was fixed by a special tool at the unit center point (UCP). For overall accuracy the film was mounted inside a phantom at a target position given by a two-dimensional cross. Its position was determined by CT or MRI scans, a treatment was planned to hit this target by use of the standard planning software and the radiation was finally delivered. This procedure is named "system test" according to DIN 6875-1 and is equivalent to a treatment simulation. The used GafChromic films were evaluated by high resolution densitometric measurements. The Munich Gamma Knife UCP coincided within x; y; z: -0.014 +/- 0.09 mm; 0.013 +/- 0.09 mm; -0.002 +/- 0.06 mm (mean +/- SD) to the center of dose distribution. There was no trend in the measured data observed over more than ten years. All measured data were within a sphere of 0.2 mm radius. When basing the target definition in the system test on MRI scans, we obtained an overall accuracy of an irradiation position in the x direction of 0.21 +/- 0.32 mm and in the y direction 0.15 +/- 0.26 mm (mean +/- SD). When a CT-based target definition was used, we measured distances in x direction 0.06 +/- 0.09 mm and in y direction 0.04 +/- 0.09 mm (mean +/- SD), respectively. These results were compared with those obtained with a Gamma Knife equipped with an automatic positioning system (APS) by use of a different phantom. This phantom was found to be slightly less accurate due to its mechanical construction and the soft fixation into the frame. The phantom related position deviation was found to be about +/- 0.2 mm, and therefore the measured accuracy of the APS Gamma Knife was evidently less precise by additional +/- 0.2 mm. These measurements demonstrate that an irradiation position defined by a CT scan can be hit within the intrinsic system precision. In radiosurgery with the Gamma Knife, a fixation with the Leksell stereotactic frame is applied. As this frame is considered to add no further uncertainties due to patient movements, the measured accuracy applies to a real patient treatment situation. The major contribution to the overall accuracy of an irradiation position is given by the MRI scans.     

GAFChromic film dosimetry with a flatbed color scanner for Leksell Gamma Knife therapy

GAFChromic films MD-55-2 have recently been established widely in industrial, scientific, and medical applications as radiation dosimeters. We applied these films to the dosimetry for Leksell Gamma Knife therapy. We used a flatbed image scanner to take digital images of irradiated MD-55-2, and the data were converted to Red, Green and Blue pixel values. The absorbed dose, as derived from the response curve of the Red pixel value, was consistent with the Leksell Gamma Plan dose planning system, for exposures using collimator sizes, 4 mm, 8 mm, and 14 mm. However, the maximum dose in the exposure of the 18 mm collimator was measured to be about 5% smaller than Gamma Plan due to the density effect of the compound material in the head phantom.     

Study of scattered photons from the collimator system of Leksell Gamma Knife using the EGS4 Monte Carlo code

In the algorithm of Leksell GAMMAPLAN (the treatment planning software of Leksell Gamma Knife), scattered photons from the collimator system are presumed to have negligible effects on the Gamma Knife dosimetry. In this study, we used the EGS4 Monte Carlo (MC) technique to study the scattered photons coming out of the single beam channel of Leksell Gamma Knife. The PRESTA (Parameter Reduced Electron-Step Transport Algorithm) version of the EGS4 (Electron Gamma Shower version 4) MC computer code was employed. We simulated the single beam channel of Leksell Gamma Knife with the full geometry. Primary photons were sampled from within the 60Co source and radiated isotropically in a solid angle of 4pi. The percentages of scattered photons within all photons reaching the phantom space using different collimators were calculated with an average value of 15%. However, this significant amount of scattered photons contributes negligible effects to single beam dose profiles for different collimators. Output spectra were calculated for the four different collimators. To increase the efficiency of simulation by decreasing the semiaperture angle of the beam channel or the solid angle of the initial directions of primary photons will underestimate the scattered component of the photon fluence. The generated backscattered photons from within the 60Co source and the beam channel also contribute to the output spectra.     

Morphology-guided radiosurgery treatment planning and optimization for multiple isocenters.

This work merges two distinct fields, 3D morphology and ionizing radiation dosimetry, to solve the problem of 3D-treatment planning and optimization in stereotactic radiosurgery. In Leksell Gamma Knife radiosurgery, dose delivery is based on the unit "shot," a dose distribution approximately spherical in shape. Multiple shots, or isocenters, are used in Gamma Knife treatment to deliver a conformal dose to an irregular radiosurgical target. The medial axis transformation, or skeleton, of the target, which uniquely characterizes the target volume and shape, is used to determine the optimal shot positions (isocenters), sizes (collimator helmet size and dosimetric weight), and the total number of shots that will deliver a conformal dose distribution to the target. The skeletonization approach reduces a complicated 3D-optimization problem to 1D searching with potential savings in computation time and mathematical complexity. In addition, optimization based on target shape replicates and automates manual treatment planning. This approach makes the process easily understandable. The relationship between skeleton discs and the dose distributions they predict is discussed. Results of optimal plans and corresponding dose distributions are presented. This approach is generally applicable to other types of multi-isocentric stereotactic radiosurgery techniques.     

A formalism for independent checking of Gamma Knife dose calculations

For stereotactic radiosurgery using the Leksell Gamma Knife system, it is important to perform a pre-treatment verification of the maximum dose calculated with the Leksell GammaPlan (DLGP) stereotactic radiosurgery system. This verification can be incorporated as part of a routine quality assurance (QA) procedure to minimize the chance of a hazardous overdose. To implement this procedure, a formalism has been developed to calculate the dose DCAL(X,Y,Z,dav,t) using the following parameters: average target depth (dav), coordinates (X,Y,Z) of the maximum dose location or any other dose point(s) to be verified, 3-dimensional (3-dim) beam profiles or off-centerratios (OCR) of the four helmets, helmet size i, output factor Oi, plug factor Pi, each shot j coordinates (x,y,z)i,j, and shot treatment time (ti,j). The average depth of the target dav was obtained either from MRI/CT images or ruler measurements of the Gamma Knife Bubble Head Frame. DCAL and DLGP were then compared to evaluate the accuracy of this independent calculation. The proposed calculation for an independent check of DLGP has been demonstrated to be accurate and reliable, and thus serves as a QA tool for Gamma Knife stereotactic radiosurgery.     

Monte Carlo calculations and GafChromic film measurements for plugged collimator helmets of Leksell Gamma Knife unit.

The Monte Carlo technique and GafChromic films were employed to verify the accuracy of the dose planning system (Leksell GammaPlan) used in Gamma Knife (type B) radiosurgery when plugged collimator helmets were used. The EGS4 Monte Carlo code was used to calculate the dose distribution along the x, y, and z axes when a single shot was delivered at the center point (unit center point: x = 100, y = 100, z = 100) of a spherical polystyrene phantom, with gamma angle of 90 degrees. Two different sizes of the plugged collimator helmets, 4 and 18 mm, were studied. Two typical plugged patterns, 51 plugs and 99 plugs along the y direction, were examined. The results of our Monte Carlo trials showed good consistency with GammaPlan calculations and GafChromic film measurements. Furthermore, the Monte Carlo results showed that radiation leakage from the plugs was too small to affect the overall isodose curve distribution even when the heavily plugged pattern of up to 99 plugs was employed. The results of this project provide confidence to all Gamma Knife centers using the Leksell GammaPlan treatment planning system.     

Monte carlo simulation of the Leksell Gamma Knife: II. Effects of heterogeneous versus homogeneous media for stereotactic radiosurgery

The absence of electronic equilibrium in the vicinity of bone-tissue or air-tissue heterogeneity in the head can misrepresent deposited dose with treatment planning algorithms that assume all treatment volume as homogeneous media. In this paper, Monte Carlo simulation (PENELOPE) and measurements with a specially designed heterogeneous phantom were applied to investigate the effect of air-tissue and bone-tissue heterogeneity on dose perturbation with the Leksell Gamma Knife. The dose fall-off near the air-tissue interface caused by secondary electron disequilibrium leads to overestimation of dose by the vendor supplied treatment planning software (GammaPlan) at up to 4 mm from an interface. The dose delivered to the target area away from an air-tissue interface may be underestimated by up to 7% by GammaPlan due to overestimation of attenuation of photon beams passing through air cavities. While the underdosing near the air-tissue interface cannot be eliminated with any plug pattern, the overdosage due to under-attenuation of the photon beams in air cavities can be eliminated by plugging the sources whose beams intersect the air cavity. Little perturbation was observed next to bone-tissue interfaces. Monte Carlo results were confirmed by measurements. This study shows that the employed Monte Carlo treatment planning is more accurate for precise dosimetry of stereotactic radiosurgery with the Leksell Gamma Knife for targets in the vicinity of air-filled cavities.     

Dosimetry of Gamma Knife and linac-based radiosurgery using radiochromic and diode detectors

In stereotactic radiosurgery the choice of appropriate detectors, whether for absolute or relative dosimetry, is very important due to the steep dose gradient and the incomplete lateral electronic equilibrium. For both linac-based and Leksell Gamma Knife radiosurgery units, we tested the use of calibrated radiochromic film to measure absolute doses and relative dose distributions. In addition a small diode was used to estimate the relative output factors. The data obtained using radiochromic and diode detectors were compared with measurements performed with other conventional methods of dosimetry, with calculated values by treatment planning systems and with data prestored in the treatment planning system supplied by the Leksell Gamma Knife (LGK) vendor. Two stereotactic radiosurgery techniques were considered: Leksell Gamma Knife (using gamma-rays from 60Co) and linac-based radiosurgery (LR) (6 MV x-rays). Different detectors were used for both relative and absolute dosimetry: relative output factors (OFs) were estimated by using radiochromic and radiographic films and a small diode; relative dose distributions in the axial and coronal planes of a spherical polystyrene phantom were measured using radiochromic film and calculated by two different treatment planning systems (TPSs). The absolute dose at the sphere centre was measured by radiochromic film and a small ionization chamber. An accurate selection of radiochromic film was made: samples of unexposed film showing a percentage standard deviation of less than 3% were used for relative dose profiles, and for absolute dose and OF evaluations this value was reduced to 1.5%. Moreover a proper calibration curve was made for each set of measurements. With regard to absolute doses, the results obtained with the ionization chamber are in good correlation with radiochromic film-generated data, for both LGK and LR, showing a dose difference of less than 1%. The output factor evaluations, performed using different methods, are in good agreement with a maximum difference of 1.5% for all field sizes considered (LGK and LR) except the 4 mm helmet used in the LGK unit. In this case, differences exist between diode and radiochromic film measurements and both detectors show data values larger than the prestored OF value of 0.80. Dose profiles measured by radiochromic film and calculated are in excellent agreement for both LGK and LR with a maximum deviation of less than 1.0 mm, when full widths of the dose profiles at 20%, 50%, 80% levels are considered. When external photon beams are used in stereotactic radiosurgery, the 'well selected' radiochromic films are very accurate detectors both for relative and absolute dosimetry. The experimental results, obtained using both radiochromic and diode detectors, show that the 4 mm helmet relative output factor could be underestimated.     

Quality assurance in stereotactic space. A system test for verifying the accuracy of aim in radiosurgery

A detailed quality assurance (QA) program is essential for high precision single dose irradiations. The accuracy of stereotactic radiosurgery is limited by the errors of each step in the chain for optimal treatment beginning with the diagnostic imaging and target localization leading to the dose planning and ending up with the treatment of the patient. Two main goals were followed on the way to finding a concept for a suitable and sufficient quality assurance routine. First, the chain of items in terms of a complete patient simulation should be followed and second, the stereotactic MR image data should be verified against a reference in our case stereotactic radiographic projection images. Target point verifications were performed using the so-called, unknown target method based on MRI, CT, and stereotactic projection images. A marked radiochromic film, embedded between inserts of the phantom is fixed parallel to either the xy or the xz plane of the stereotactic coordinate system. After imaging and planning, the phantom is adjusted and irradiated. At the end, the film, dyed by the radiation field around the premarked cross, is evaluated. The measured distance between the unit center point (shadow) and the localization of the marked film leads to the deviation to be minimized. This is referred to as the displacement vector. The results, evaluating 170 system tests within 5 years. show that the mean displacement vector of the complete system is 0.48 mm +/-0.23 mm (mean+/- sd). Factors having a significant influence on the overall accuracy are associated with MRI parameters. Test results based on axial images (xy plane; 0.42 mm +/- 0.24 mm) are significantly superior to coronal images (xz plane; x = 0.60 mm +/- 0.02 mm). Further on, the 3D-mpr sequence (0.40 mm +/- 0.19 mm) is significantly superior to the T1 weighted SE sequences (0.66 mm +/- 0.24 mm). Given the high mechanical accuracy of the Leksell gamma knife, the most sensitive technical factor having an influence on the overall precision of radiosurgery is the MRI study. However, using the appropriate imaging sequences and parameters the dislocation error inferred by MRI can be kept very low and restricted to the rare patient inherent distortion factors. With these precautions in mind, MRI is recommended as the imaging method of choice in radiosurgery.     

A simplified model of the source channel of the Leksell Gamma Knife: testing multisource configurations with PENELOPE

A simplification of the source channel geometry of the Leksell Gamma Knife (GK), recently proposed by the authors and checked for a single source configuration (Al-Dweri F M O, Lallena A M and Vilches M 2004 Phys. Med. Biol. 49 2687-703), has been used to calculate the dose distributions along the x, y and z axes in a water phantom with a diameter of 160 mm, for different configurations of the Gamma Knife, including 201, 150 and 102 unplugged sources. The code PENELOPE (v. 2001) has been used to perform the Monte Carlo simulations. In addition, the output factors for the 14, 8 and 4 mm helmets have been calculated. The results found for the dose profiles show a qualitatively good agreement with previous ones obtained with EGS4 and PENELOPE (v. 2000) codes and with the predictions of GammaPlan. The output factors obtained with our model agree within the statistical uncertainties with those calculated with the same Monte Carlo codes and with those measured with different techniques. Owing to the accuracy of the results obtained and to the reduction in the computational time with respect to full geometry simulations (larger than a factor 15), this simplified model opens the possibility of using Monte Carlo tools for planning purposes in the Gamma Knife.     

Gamma Knife 3-D dose distribution near the area of tissue inhomogeneities by normoxic gel dosimetry

The accuracy of the Leksell GammaPlan, the dose planning system of the Gamma Knife Model-B, was evaluated near tissue inhomogeneities, using the gel dosimetry method. The lack of electronic equilibrium around the small-diameter gamma beams can cause dose calculation errors in the neighborhood of an air-tissue interface. An experiment was designed to investigate the effects of inhomogeneity near the paranosal sinuses cavities. The homogeneous phantom was a spherical glass balloon of 16 cm diameter, filled with MAGIC gel; i.e., the normoxic polymer gel. Two hollow PVC balls of 2 cm radius, filled with N2 gas, represented the air cavities inside the inhomogeneous phantom. For dose calibration purposes, 100 ml gel-containing vials were irradiated at predefined doses, and then scanned in a MR unit. Linearity was observed between the delivered dose and the reciprocal of the T2 relaxation time constant of the gel. Dose distributions are the results of a single shot of irradiation, obtained by collimating all 201 cobalt sources to a known target in the phantom. Both phantoms were irradiated at the same dose level at the same coordinates. Stereotactic frames and fiducial markers were attached to the phantoms prior to MR scanning. The dose distribution predicted by the Gamma Knife planning system was compared with that of the gel dosimetry. As expected, for the homogeneous phantom the isodose diameters measured by the gel dosimetry and the GammaPlan differed by 5% at most. However, with the inhomogeneous phantom, the dose maps in the axial, coronal and sagittal planes were spatially different. The diameters of the 50% isodose curves differed 43% in the X axis and 32% in the Y axis for the Z =90 mm axial plane; by 44% in the X axis and 24% in the Z axis for the Y=90 mm coronal plane; and by 32% in the Z axis and 42% in the Y axis for the X=92 mm sagittal plane. The lack of ability of the GammaPlan to predict the rapid dose fall off, due to the air cavities behind or near the lesion led to an overestimation of the dose that was actually delivered. Clinically, this can result in underdosing of lesions near tissue inhomogeneities in patients under treatment.     

Investigation of gamma knife image registration errors resulting from misalignment between the patient and the imaging axis

The ability of Leksell GammaPlan to perform stereotactic space localizations with image sets where there is misalignment of the patient's head (stereotactic frame and fiducial apparatus) relative to the computed tomography (CT) scanner coordinate system was studied. Misalignment is sometimes necessary for patient comfort. Results equally apply to magnetic resonance imaging. Seven 0.5 mm diameter CT-visible spheres were rigidly mounted to a string tied tightly at each end to diagonally opposite posts attached to a Leksell stereotactic frame. A standard CT fiducial box was applied to the frame in the usual clinical manner. A baseline CT scan (1 mm slice thickness) was obtained with the fiducial box perfectly aligned with the scanner axis. After localization of the image set, the (x,y,z) coordinate of the center of each sphere was recorded. Repeat CT scans with varying fiducial box misalignments with the imaging axis were subsequently obtained. The mean difference between the base line and the respective coordinates in misaligned geometries was approximately 0.2 mm (sigma=0.2 mm), well within the accuracy of the image sets and the delivery of radiosurgery with the Gamma Knife.     

Effects of bone- and air-tissue inhomogeneities on the dose distributions of the Leksell Gamma Knife calculated with PENELOPE

Monte Carlo simulation with PENELOPE (version 2003) is applied to calculate Leksell Gamma Knife dose distributions for heterogeneous phantoms. The usual spherical water phantom is modified with a spherical bone shell simulating the skull and an air-filled cube simulating the frontal or maxillary sinuses. Different simulations of the 201 source configuration of the Gamma Knife have been carried out with a simplified model of the geometry of the source channel of the Gamma Knife recently tested for both single source and multisource configurations. The dose distributions determined for heterogeneous phantoms including the bone- and/or air-tissue interfaces show non-negligible differences with respect to those calculated for a homogeneous one, mainly when the Gamma Knife isocentre approaches the separation surfaces. Our findings confirm an important underdosage (approximately 10%) nearby the air-tissue interface, in accordance with previous results obtained with the PENELOPE code with a procedure different from ours. On the other hand, the presence of the spherical shell simulating the skull produces a few per cent underdosage at the isocentre wherever it is situated.     

Assessment of absorbed dose to thyroid, parotid and ovaries in patients undergoing Gamma Knife radiosurgery

Stereotactic radiosurgery was originally introduced by Lars Leksell in 1951. This treatment refers to the noninvasive destruction of an intracranial target localized stereotactically. The purpose of this study was to identify the dose delivered to the parotid, ovaries, testis and thyroid glands during the Gamma Knife radiosurgery procedure. A three-dimensional, anthropomorphic phantom was developed using natural human bone, paraffin and sodium chloride as the equivalent tissue. The phantom consisted of a thorax, head and neck and hip. In the natural places of the thyroid, parotid (bilateral sides) and ovaries (midline), some cavities were made to place TLDs. Three TLDs were inserted in a batch with 1 cm space between the TLDs and each batch was inserted into a single cavity. The final depth of TLDs was 3 cm from the surface for parotid and thyroid and was 15 cm for the ovaries. Similar batches were placed superficially on the phantom. The phantom was gamma irradiated using a Leksell model C Gamma Knife unit. Subsequently, the same batches were placed superficially over the thyroid, parotid, testis and ovaries in 30 patients (15 men and 15 women) who were undergoing radiosurgery treatment for brain tumours. The mean dosage for treating these patients was 14.48 +/- 3.06 Gy (10.5-24 Gy) to a mean tumour volume of 12.30 +/- 9.66 cc (0.27-42.4 cc) in the 50% isodose curve. There was no significant difference between the superficial and deep batches in the phantom studies (P-value < 0.05). The mean delivered doses to the parotid, thyroid, ovaries and testis in human subjects were 21.6 +/- 15.1 cGy, 9.15 +/- 3.89 cGy, 0.47 +/- 0.3 cGy and 0.53 +/- 0.31 cGy, respectively. The data can be used in making decisions for special clinical situations such as treating pregnant patients or young patients with benign lesions who need radiosurgery for eradication of brain tumours.