Clinical dosimetry for implementation of a multileaf collimator.

In order to initiate the use of a multileaf collimator (MLC) in the clinic, a set of technical procedures needs to be available sufficient to create MLC leaf settings and to deliver an accurate dose of radiation through the MLC-shaped field. Dosimetry data for clinical use of the MLC were measured. Dosimetric characteristics included central axis percent depth dose, output factors, and penumbra. In this paper, it has been concluded that a dose control monitor unit calculation procedure that has been applied to the use of conventional secondary field-shaping blocks can be applied to the multileaf collimator dosimetry. The multileaf collimator penumbra (20% to 80%) is only slightly wider (1-3 mm) than the penumbra of the conventional collimator jaws. Beam's-eye-view comparisons made between the isodose curves in fields shaped by conventional Cerrobend blocks and isodose curves in fields shaped by the multileaf collimator demonstrated that the 50% isodose line at 10-cm depth exhibited the discrete steps of the multileaf collimator leaves, but that the 90% and 10% isodose curves of the multileaf were close to those shaped by Cerrobend blocks.     

Improving IMRT quality control efficiency using an amorphous silicon electronic portal imager

An amorphous silicon electronic portal imaging device (EPID) has been investigated to determine its usefulness and efficiency for performing linear accelerator quality control checks specific to step and shoot intensity modulated radiation therapy (IMRT). Several dosimetric parameters were measured using the EPID: dose linearity and segment to segment reproducibility of low dose segments, and delivery accuracy of fractions of monitor units. Results were compared to ion chamber measurements. Low dose beam flatness and symmetry were tested by overlaying low dose beam profiles onto the profile from a stable high-dose exposure and visually checking for differences. Beam flatness and symmetry were also calculated and plotted against dose. Start-up reproducibility was tested by overlaying profiles from twenty successive two monitor unit segments. A method for checking the MLC leaf calibration was also tested, designed to be used on a daily or weekly basis, which consisted of summing the images from a series of matched fields. Daily images were coregistered with, then subtracted from, a reference image. A threshold image showing dose differences corresponding to > 0.5 mm positional errors was generated and the number of pixels with such dose differences used as numerical parameter to which a tolerance can be applied. The EPID was found to be a sensitive relative dosemeter, able to resolve dose differences of 0.01 cGy. However, at low absolute doses a reproducible dosimetric nonlinearity of up to 7% due to image lag/ghosting effects was measured. It was concluded that although the EPID is suitable to measure segment to segment reproducibility and fractional monitor unit delivery accuracy, it is still less useful than an ion chamber as a tool for dosimetric checks. The symmetry/flatness test proved to be an efficient method of checking low dose profiles, much faster than any of the alternative methods. The MLC test was found to be extremely sensitive to sudden changes in MLC calibration but works best with a composite reference image consisting of an average of five successive days' images. When used in this way it proved an effective and efficient daily check of MLC calibration. Overall, the amorphous silicon EPID was found to be a suitable device for IMRT QC although it is not recommended for dosimetric tests. Automatic procedures for low monitor unit profile analysis and MLC leaf positioning yield considerable time-savings over traditional film techniques.     

A dosimetric comparison of two multileaf collimator designs

We present the results of measurements designed to compare two different multileaf collimator (MLC) designs using a novel evaluation technique. The MLC designs evaluated were: a "single-focused" MLC (SF-MLC) mounted below the jaws, and a "double-focused" MLC, which is a complete replacement for the lower jaws. The ability of each MLC to conform isodose lines to a prescribed field edge (PFE) was evaluated using film dosimetry. Circular fields, centered on axis and off axis, were used because they produce a range of "angles of approach" between the MLC leaves and the PFE. They also have the advantage that for an ideal field shaping system the resulting isodoses are concentric perfect circles, a well-defined basis for evaluation. The amplitude of the oscillations of the 50% isodose line about the PFE and the penumbra width as determined by the 20%, 80%, and 90% isodose lines was examined. We observe that the 50% isodose line oscillates around the PFE with greater amplitude for SF-MLC. We attribute this, at least in part, to the rounded ends of the SF-MLC leaves. However, the SF-MLC has a noticeably sharper penumbra, which we attribute to its position further from the source. We conclude that these results are relevant for accurate dosimetric modeling of these devices.     

Dose distributions of x-ray fields as shaped with multileaf collimators.

Multileaf collimators (MLC) with various blade widths were simulated using standard cerrobend blocks, and three-dimensional dose computations were carried out to study the resultant radiation field edges. Film measurements made with 6 and 18 MV x-ray beams were compared with calculations that employed a three-dimensional Fourier convolution. A spatial accuracy of better than 3 mm was found in the 50% isodose line of the penumbral region with a calculation voxel size of 5 mm x 5 mm x 5 mm. The computer simulation was used to study the deviation of the calculated 50% isodose line from the desired geometric field edge using various MLC blade positions. The study suggests that multileaf collimation to the outside of the desired field edge will lead to overdose outside the field, whereas multileaf collimation to the inside of the desired field edge will lead to underdose inside the field. When the direction of travel of the leaves with respect to the field edge is near 45 degrees, the 50% isodose of a multileaf-collimated beam will fall close to the desired edge with no underdose when the leaf corners are allowed to insert into the desired field edge by 1.2 mm for 6 MV x-rays and 1.4 mm for 18 MV x-rays using a 1 cm wide leaf. These blade offsets account for the scattering of photons and electrons in the medium within the penumbral region.     

A dose verification method using a monitor unit matrix for dynamic IMRT on Varian linear accelerators

Dosimetry verification is an important step during intensity modulated radiotherapy treatment (IMRT). The verification is usually conducted with measurements and independent dose calculations. However, currently available independent dose calculation methods were developed for step-and-shoot beam delivery methods, and their uses for dynamic multi-leaf collimator (MLC) delivery methods are not efficient. In this study, a dose calculation method was developed to perform independent dose verifications for a dynamic MLC-based IMRT technique for Varian linear accelerators. This method extracts the machine delivery parameters from the dynamic MLC (DMLC) files generated by the IMRT treatment planning system. Based on the parameters a monitor unit (MU) matrix was separately calculated as two terms: direct exposure from the open MLC field and leakage contributions, where the leaf-end leakage contribution becomes more important in higher dose gradient regions. The MU matrix was used to compute the primary dose and the scattered dose with a modified Clarkson technique. The doses computed using the method were compared with both measurement and treatment planning for 14 and 25 plans respectively. An average of less than 2% agreement was observed and the standard deviation was about 1.9%.     

Validation of dynamic MLC-controller log files using a two-dimensional diode array

Intensity-modulated radiation therapy (IMRT) delivered with multi-leaf collimator (MLC) in the step-and-shoot mode uses multiple static MLC segments to achieve intensity modulation. For typical IMRT treatment plans, significant numbers of segments are delivered with monitor units (MUs) of much less than 10. Verification of the ability of the linear accelerator (linac) to deliver small MU segments accurately is an important step in the IMRT commissioning and quality assurance (QA) process. Recent studies have reported large discrepancies between the intended and delivered segment MUs. These discrepancies could potentially cause large errors in the delivered patient dose. We have undertaken a systematic study to evaluate the accuracy of the dynamic MLC log files, which are created automatically by our commercial MLC workstation after each delivery, in recording the fractional MU delivered in the step-and-shoot mode. Two linac models were evaluated with simple-geometry leaf sequences and delivered with different total MUs and different nominal dose rates. A commercial two-dimensional diode array was used for the measurement. Large discrepancies between the intended and delivered segment MUs were found. The discrepancies were larger for small MU segments at higher dose rate, with some small MU segments completely undelivered. The recorded fractional MUs in the log files were found to agree with what was delivered within the limits of our experimental uncertainty. Our results indicate that it is important to verify the delivery accuracy of small MU segments that could potentially occur in a patient treatment and that the log files are useful in checking the integrity of the linac delivery once validated. Thus validated log files can be used as a QA tool for general IMRT delivery and patient-specific plan verification.     

An independent dose calculation algorithm for MLC-based radiotherapy including the spatial dependence of MLC transmission

An analytical dose calculation algorithm was developed and commissioned to calculate dose delivered with both static and dynamic multileaf collimator (MLC) in a homogenous phantom. The algorithm is general; however, it was designed specifically to accurately model dose for large and complex IMRT fields. For such fields the delivered dose may have a considerable contribution from MLC transmission, which is dependent upon spatial considerations. Specifically, the algorithm models different MLC effects, such as interleaf transmission, the tongue-and-groove effect, rounded leaf ends, MLC scatter, beam hardening and divergence of the beam, which results in a gradual MLC transmission fall-off with increasing off-axis distance. The calculated dose distributions were compared to measured dose using different methods (film, ionization chamber array, single ionization chamber), and the differences among the treatment planning system, the measurements and the developed algorithm were analysed for static MLC and dynamic IMRT fields. It was found that the calculated dose from the developed algorithm agrees very well with the measurements (mostly within 1.5%) and that a constant value for MLC transmission is insufficient to accurately predict dose for large targets and complex IMRT plans with many monitor units.     

An empirical model for independent dose verification of the Gamma Knife treatment planning

A formalism for an independent dose verification of the Gamma Knife treatment planning is developed. It is based on the approximation that isodose distribution for a single shot is in the shape of an ellipsoid in three-dimensional space. The dose profiles for a phantom along each of the three major axes are fitted to a function which contains the terms that represent the contributions from a point source, an extrafocal scattering, and a flat background. The fitting parameters are extracted for all four helmet collimators, at various shot locations, and with different skull shapes. The 33 parameters of a patient's skull shape obtained from the Skull Scaling Instrument measurements are modeled for individual patients. The relative doses for a treatment volume in the form of 31 x 31 x 31 matrix of points are extracted from the treatment planning system, the Leksell Gamma-Plan (LGP). Our model evaluates the relative doses using the same input parameters as in the LGP, which are skull measurement data, shot location, weight, gamma-angle of the head frame, and helmet collimator size. For 29 single-shot cases, the discrepancy of dose at the focus point between the calculation and the LGP is found to be within -1% to 2%. For multi-shot cases, the value and the coordinate of the maximum dose point from the calculation agree within +/-7% and +/-3 mm with the LGP results. In general, the calculated doses agree with the LGP calculations within +/-10% for the off-center locations. Results of calculation with this method for the dimension and location of the 50% isodose line are in good agreement with results from Leksell GammaPlan. Therefore, this method can be served as a useful tool for secondary quality assurance of Gamma Knife treatment plans.     

A graph-searching method for MLC leaf sequencing under constraints

A new leaf-sequencing algorithm for step-and-shoot IMRT that is based on a graph-searching technique is described. An iterative process guided by a quantitative measure for the complexity of the initial or residual intensity pattern is used to identify the field segments shaped by a multileaf collimator (MLC). Given a user selected number of intensity levels, the algorithm searches deliverable segment candidates considering all intensity levels and two collimator positions separated by 90 degrees. The candidates for each intensity level are obtained as the least number of segments to cover the areas with equal or higher intensity. The shape of a deliverable segment is adjusted by leaving out certain beam elements for later delivery if this results in a simpler residual intensity pattern and the segment is still deliverable. For a MLC design that does not allow leaf interdigitation, it is initially assumed that a single segment cannot cover two disjoined areas. Among all candidates the segment with the greatest reduction of the complexity of the residual intensity distribution is chosen for the current step of iteration. The iterative process generates a set of deliverable segments of simply connected areas. These segments are combined later under specific MLC constraints. Different orders of segment combination are considered for minimizing the beam-on time. The final segments are sequenced to minimize the leaf travel. This algorithm has been tested using randomly generated intensity distributions and clinical cases for the Varian, Siemens, and Elekta MLC systems. The results show that as the number of intensity levels is increased, the numbers of segments and MUs increase only modestly. Using two collimator angles results in decreases in the required number of segments and the number of monitor units that can be as much as 20%.     

Dosimetric verification of a commercial inverse treatment planning system

A commercial three-dimensional (3D) inverse treatment planning system, Corvus (Nomos Corporation, Sewickley, PA), was recently made available. This paper reports our preliminary results and experience with commissioning this system for clinical implementation. This system uses a simulated annealing inverse planning algorithm to calculate intensity-modulated fields. The intensity-modulated fields are divided into beam profiles that can be delivered by means of a sequence of leaf settings by a multileaf collimator (MLC). The treatments are delivered using a computer-controlled MLC. To test the dose calculation algorithm used by the Corvus software, the dose distributions for single rectangularly shaped fields were compared with water phantom scan data. The dose distributions predicted to be delivered by multiple fields were measured using an ion chamber that could be positioned in a rotatable cylindrical water phantom. Integrated charge collected by the ion chamber was used to check the absolute dose of single- and multifield intensity modulated treatments at various spatial points. The measured and predicted doses were found to agree to within 4% at all measurement points. Another set of measurements used a cubic polystyrene phantom with radiographic film to record the radiation dose distribution. The films were calibrated and scanned to yield two-dimensional isodose distributions. Finally, a beam imaging system (BIS) was used to measure the intensity-modulated x-ray beam patterns in the beam's-eye view. The BIS-measured images were then compared with a theoretical calculation based on the MLC leaf sequence files to verify that the treatment would be executed accurately and without machine faults. Excellent correlation (correlation coefficients > or = 0.96) was found for all cases. Treatment plans generated using intensity-modulated beams appear to be suitable for treatment of irregularly shaped tumours adjacent to critical structures. The results indicated that the system has potential for clinical radiation treatment planning and delivery and may in the future reduce treatment complexity.     

Verification of IMRT fields by film dosimetry

In intensity modulated radiation therapy (IMRT) the aim of an accurate conformal dose distribution is obtained through a complex process. This ranges from the calculation of the optimal distribution of fluence by the treatment planning system (TPS), to the dose delivery through a multilamellar collimator (MLC), with several segments per beam in the step and shoot approach. The above-mentioned consideration makes mandatory an accurate dosimetric verification of the IM beams. A high resolution and integrating dosimeter, like the radiographic film, permits one to simultaneously measure the dose in a matrix of points, providing a good means of obtaining dose distributions. The intrinsic limitation of film dosimetry is the sensitivity dependence on the field size and on the measurement depth. However, the introduction of a scattered radiation filter permits the use of a single calibration curve for all field sizes and measurement depths. In this paper the quality control procedure developed for dosimetric verification of IMRT technique is reported. In particular a system of film dosimetry for the verification of a 6 MV photon beam has been implemented, with the introduction of the scattered radiation filter in the clinical practice that permits one to achieve an absolute dose determination with a global uncertainty within 3.4% (1 s.d.). The film has been calibrated to be used both in perpendicular and parallel configurations. The work also includes the characterization of the Elekta MLC. Ionimetric independent detectors have been used to check single point doses. The film dosimetry procedure has been applied to compare the measured absolute dose distributions with the ones calculated by the TPS, both for test and clinical plans. The agreement, quantified by the gamma index that seldom reaches the 1.5 value, is satisfying considering that the comparison is performed between absolute doses.     

An accurate calibration method of the multileaf collimator valid for conformal and intensity modulated radiation treatments

Because for IMRT treatments the required accuracy on leaf positioning is high, conventional calibration methods may not be appropriate. The aim of this study was to develop the tools for an accurate MLC calibration valid for conventional and IMRT treatments and to investigate the stability of the MLC. A strip test consisting of nine adjacent segments 2 cm wide, separated by 1 mm and exposed on Kodak X-Omat V films at Dmax depth, was used for detecting leaf-positioning errors. Dose profiles along the leaf-axis were taken for each leaf-pair. We measured the dose variation on each abutment to quantify the relative positioning error (RPE) and the absolute position of the abutment to quantify the absolute positioning error (APE). The accuracy of determining the APE and RPE was 0.15 and 0.04 mm, respectively. Using the RPE and the APE the MLC calibration parameters were calculated in order to obtain a flat profile on the abutment at the correct position. A conventionally calibrated Elekta MLC was re-calibrated using the strip test. The stability of the MLC and leaf-positioning reproducibility was investigated exposing films with 25 adjacent segments 1 cm wide during three months and measuring the standard deviation of the RPE values. A maximum shift over the three months of 0.27 mm was observed and the standard deviation of the RPE values was 0.11 mm.     

Investigation of secondary neutron dose for 18 MV dynamic MLC IMRT delivery

Secondary neutron doses from the delivery of 18 MV conventional and intensity modulated radiation therapy (IMRT) treatment plans were compared. IMRT was delivered using dynamic multileaf collimation (MLC). Additional measurements were made with static MLC using a primary collimated field size of 10 x 10 cm2 and MLC field sizes of 0 x 0, 5 x 5, and 10 x 10 cm2. Neutron spectra were measured and effective doses calculated. The IMRT treatment resulted in a higher neutron fluence and higher dose equivalent. These increases were approximately the ratio of the monitor units. The static MLC measurements were compared to Monte Carlo calculations. The actual component dimensions and materials for the Varian Clinac 2100/2300C including the MLC were modeled with MCNPX to compute the neutron fluence due to neutron production in and around the treatment head. There is excellent agreement between the calculated and measured neutron fluence for the collimated field size of 10 x 10 cm2 with the 0 x 0 cm2 MLC field. Most of the neutrons at the detector location for this geometry are directly from the accelerator head with a small contribution from room scatter. Future studies are needed to investigate the effect of different beam energies used in IMRT incorporating the effects of scattered photon dose as well as secondary neutron dose.     

Computation of dose distribution for linear radioactive sources in brachytherapy.

This paper describes a microcomputer program, written in Quick BASIC (Version 4.5), that allows the planning of radio-therapeutic doses in brachytherapy. The program allows the calculation of treatment times of prescribed doses, dose rates in selected points, as well as the computation of 2D and 3D isodose contours, of any linear source distribution. Also the volume enclosed by a chosen dose rate can be calculated. The results agree better than 1% with the results reported in the literature for discrete dose rate values, and for isodose contours.     

Segmental and dynamic intensity-modulated radiotherapy delivery techniques for micro-multileaf collimator

A leaf sequencing algorithm has been implemented to deliver segmental and dynamic multileaf collimated intensity-modulated radiotherapy (SMLC-IMRT and DMLC-IMRT, respectively) using a linear accelerator equipped with a micro-multileaf collimator (mMLC). The implementation extends a previously published algorithm for the SMLC-IMRT to include the dynamic MLC-IMRT method and several dosimetric considerations. The algorithm has been extended to account for the transmitted radiation and minimize the leakage between opposing and neighboring leaves. The underdosage problem associated with the tongue-and-groove design of the MLC is significantly reduced by synchronizing the MLC leaf movements. The workings of the leaf sequencing parameters have been investigated and the results of the planar dosimetric investigations show that the sequencing parameters affect the measured dose distributions as intended. Investigations of clinical cases suggest that SMLC and DMLC delivery methods produce comparable results with leaf sequences obtained by root-mean-square (RMS) errors specification of 1.5% and lower, approximately corresponding to 20 or more segments. For SMLC-IMRT, there is little to be gained by using an RMS error specification smaller than 2%, approximately corresponding to 15 segments; however, more segments directly translate to longer treatment time and more strain on the MLC. The implemented leaf synchronization method does not increase the required monitor units while it reduces the measured TG underdoses from a maximum of 12% to a maximum of 3% observed with single field measurements of representative clinical cases studied.     

A dosimetric leaf-setting strategy for shaping radiation fields using a multileaf collimator

A dosimetric leaf-setting strategy of using multileaf collimators (MLC) for shaping radiation fields has been developed. Existing MLC leaf-setting strategies are all based upon geometric criteria. This new approach, however, matches a prescribed field contour with a MLC using clinically consistent dosimetric criteria. The leaf positions are determined using an iterative optimization algorithm. An empirical dose model was developed to compare the dosimetric-based leaf-setting strategy with the geometric-based leaf-setting strategies. Differences up to half a centimeter in the leaf positions and isodose lines were found between setting the MLC geometrically and setting the MLC dosimetrically. The dosimetric leaf-setting strategy provides the ability to achieve better dose conformation for a clinically desired isodose line. Since the desired isodose line that covers a treatment volume is typically higher than 50% of the maximum dose, the scalloping effects due to the finite leaf width at the leaf edge or 50% isodose lines are much reduced. Another benefit of the dosimetric leaf-setting is that it separates the leaf-setting process from the treatment planning process, and this frees the treatment planning vendors from developing detailed dose models for various existing types and future upgrades of MLC systems.     

The reproducibility of polyacrylamide gel dosimetry applied to stereotactic conformal radiotherapy

The reproducibility of polyacrylamide gel (PAG) dosimetry has been evaluated when used to verify two radiotherapy treatment plans of increasing complexity. The plans investigated were a three-field coplanar arrangement, using the linac jaws for field shaping, and a four-field, conformal, non-coplanar plan using precision-cast lead alloy shielding blocks. Each treatment was performed three times using phantoms and calibration gels manufactured in-house. Two phantoms were specially designed for this work to aid accurate positioning of the gels for irradiation and imaging. All gels were imaged post-irradiation using a Siemens Vision 1.5T MR scanner. T2 relaxation images were calibrated to absorbed dose distributions using a number of smaller calibration vessels to produce distribution maps of relative dose. The relative dose distributions were found to be reproducible, with the standard deviation on the mean areas enclosed by the > or = 50% isodose lines measured in three orthogonal planes being 6.4% and 4.1% for the coplanar and non-coplanar plans respectively. The measured distributions were also consistent with those planned, with isodose lines generally agreeing to within a few millimetres. However, the measured absolute doses were on average 23.5% higher than those planned. Although the polyacrylamide gel dosimetry technique has some limitations, particularly when calibrating distributions to absolute dose, the ability to resolve sharp dose gradients in three dimensions with millimetre precision is invaluable when verifying complex conformal treatment plans, where avoidance of proximal, critical structures is a treatment criterion.     

Independent dosimetric calculation with inclusion of head scatter and MLC transmission for IMRT

Independent verification of the MU settings and dose calculation of IMRT treatment plans is an important step in the IMRT quality assurance (QA) procedure. At present, the verification is mainly based on experimental measurements, which are time consuming and labor intensive. Although a few simplified algorithms have recently been proposed for the independent dose (or MU) calculation, head scatter has not been precisely taken into account in all these investigations and the dose validation has mainly been limited to the central axis. In this work we developed an effective computer algorithm for IMRT MU and dose validation. The technique is superior to the currently available computer-based MU check systems in that (1) it takes full consideration of the head scatter and leaf transmission effects; and (2) it allows a precise dose calculation at an arbitrary spatial point instead of merely a point on the central axis. In the algorithm the dose at an arbitrary spatial point is expressed as a summation of the contributions of primary and scatter radiation from all beamlets. Each beamlet is modulated by a dynamic modulation factor (DMF), which is determined by the MLC leaf trajectories, the head scatter, the jaw positions, and the MLC leaf transmission. A three-source model was used to calculate the head scatter distribution for irregular segments shaped by MLC and the scatter dose contributions were computed using a modified Clarkson method. The system reads in MLC leaf sequence files (or RTP files) generated by the Corvus (NOMOS Corporation, Sewickley, PA) inverse planning system and then computes the doses at the desired points. The algorithm was applied to study the dose distributions of several testing intensity modulated fields and two multifield Corvus plans and the results were compared with Corvus plans and experimental measurements. The final dose calculations at most spatial points agreed with the experimental measurements to within 3% for both the specially designed testing fields and the clinical intensity modulated field. Furthermore, excellent agreement (mostly within +/- 3.0%) was also found between our independent calculation and the ion chamber measurements at both central axis and off-axis positions for the multifield Corvus IMRT plans. These results indicate that the approach is robust and valuable for routine clinical IMRT plan validation.     

Output ratio in air for MLC shaped irregular fields

For accurate monitor unit calculation, it is important to calculate the output ratio in air, Sc, for an irregular field shaped by MLC. We have developed an algorithm to calculate Sc based on an empirical model [Med. Phys. 28, 925-937 (2001)] by projecting each leaf position to the isocenter plane. Thus it does not require the exact knowledge of the head geometry. Comparisons were made for three different types of MLC: those with MLC replacing the inner collimator jaws; those with MLC replacing the outer collimator jaws; and those with MLC as a tertiary attachment. When the MLC leaf positions are substantially different from the secondary collimators (or the rectangular field encompassing the irregular field), one observes an up to 5% difference in the value of head-scatter correction factor, HCF, defined as the ratio of output ratio in air between the MLC shaped irregular field and that of the rectangular field encompassing the irregular field. No collimator exchange effect was observed for rectangular fields shaped by MLC (e.g., 5x30 and 30x5 cm2 diagonal) when the secondary collimators are fixed, unlike that for the rectangular fields shaped by the inner and outer collimator jaws, where it can be 1-2%. For the same MLC shaped irregular field, the value of Sc increases from the Elekta, to the Siemens, to the Varian accelerators, with an up to 4% difference. The calculation agrees with measurement to within 1.2% for points both on and off the central-axis. The fitting parameters used in the algorithm are derived from measurements for square field sizes on the central-axis.     

Incorporating multi-leaf collimator leaf sequencing into iterative IMRT optimization

Intensity modulated radiation therapy (IMRT) treatment planning typically considers beam optimization and beam delivery as separate tasks. Following optimization, a multi-leaf collimator (MLC) or other beam delivery device is used to generate fluence patterns for patient treatment delivery. Due to limitations and characteristics of the MLC, the deliverable intensity distributions often differ from those produced by the optimizer, leading to differences between the delivered and the optimized doses. Objective function parameters are then adjusted empirically, and the plan is reoptimized to achieve a desired deliverable dose distribution. The resulting plan, though usually acceptable, may not be the best achievable. A method has been developed to incorporate the MLC restrictions into the optimization process. Our in-house IMRT system has been modified to include the calculation of the deliverable intensity into the optimizer. In this process, prior to dose calculation, the MLC leaf sequencer is used to convert intensities to dynamic MLC sequences, from which the deliverable intensities are then determined. All other optimization steps remain the same. To evaluate the effectiveness of deliverable-based optimization, 17 patient cases have been studied. Compared with standard optimization plus conversion to deliverable beams, deliverable-based optimization results show improved isodose coverage and a reduced dose to critical structures. Deliverable-based optimization results are close to the original nondeliverable optimization results, suggesting that IMRT can overcome the MLC limitations by adjusting individual beamlets. The use of deliverable-based optimization may reduce the need for empirical adjustment of objective function parameters and reoptimization of a plan to achieve desired results.