Mike Oliver

Segmentation and leaf sequencing for intensity modulated arc therapy

A common method in generating intensity modulated radiation therapy (IMRT) plans consists of a three step process: an optimized fluence intensity map (IM) for each beam is generated via inverse planning, this IM is then segmented into discrete levels, and finally, the segmented map is translated into a set of MLC apertures via a leaf sequencing algorithm. To date, limited work has been done on this approach as it pertains to intensity modulated arc therapy (IMAT), specifically in regards to the latter two steps. There are two determining factors that separate IMAT segmentation and leaf sequencing from their IMRT equivalents: (1) the intrinsic 3D nature of the intensity maps (standard 2D maps plus the angular component), and (2) that the dynamic multileaf collimator (MLC) constraints be met using a minimum number of arcs. In this work, we illustrate a technique to create an IMAT plan that replicates Tomotherapy deliveries by applying IMAT specific segmentation and leaf-sequencing algorithms to Tomotherapy output sinograms. We propose and compare two alternative segmentation techniques, a clustering method, and a bottom-up segmentation method (BUS). We also introduce a novel IMAT leaf-sequencing algorithm that explicitly takes leaf movement constraints into consideration. These algorithms were tested with 51 angular projections of the output leaf-open sinograms generated on the Hi-ART II treatment planning system (Tomotherapy Inc.). We present two geometric phantoms and 2 clinical scenarios as sample test cases. In each case 12 IMAT plans were created, ranging from 2 to 7 intensity levels. Half were generated using the BUS segmentation and half with the clustering method. We report on the number of arcs produced as well as differences between Tomotherapy output sinograms and segmented IMAT intensity maps. For each case one plan for each segmentation method is chosen for full Monte Carlo dose calculation (NumeriX LLC) and dose volume histograms (DVH) are calculated. In all cases, the BUS method outperformed the clustering, method. We recommend using the BUS algorithm and discuss potential improvements to the clustering algorithms.

Monte Carlo dose calculation of segmental IMRT delivery to a moving phantom using dynamic MLC and gating log files

Respiratory gating is emerging as a tool to limit the effect of motion for liver and lung tumors. In order to study the impact of target motion and gated intensity modulated radiation therapy (IMRT) delivery, a computer program was developed to simulate segmental IMRT delivery to a moving phantom. Two distinct plans were delivered to a rigid-motion phantom with a film insert in place under four conditions: static, sinusoidal motion, gated sinusoidal motion with a duty cycle of 25% and gated sinusoidal motion with duty cycle of 50% under motion conditions of a typical patient (A = 1 cm, T = 4 s). The MLC controller log files and gating log files were retained to perform a retrospective Monte Carlo dose calculation of the plans. Comparison of the 2D planar dose distributions between simulation and measurement demonstrated that our technique had at least 94% of the points passing gamma criteria of 3% for dose difference and 3 mm as the distance to agreement. This note demonstrates that the use of dynamic multi-leaf collimator and respiratory monitoring system log files together with a fast Monte Carlo dose calculation algorithm is an accurate and efficient way to study the dosimetric effect of motion for gated or non-gated IMRT delivery on a rigidly-moving body.

Experimental measurements and Monte Carlo simulations for dosimetric evaluations of intrafraction motion for gated and ungated intensity modulated arc therapy deliveries.

Respiratory gated radiation therapy allows for a smaller margin expansion for the planning target volume (PTV) to account for respiratory induced motion and is emerging as a common method to treat lung and liver tumors. We investigated the dosimetric effect of free motion and gated delivery for intensity modulated arc therapy (IMAT) with experimental measurements and Monte Carlo simulations. The impact of PTV margin and duty cycle for gated delivery is studied with Monte Carlo simulations. A motion phantom is used for this study. Two sets of contours were drawn on the mid-inspiration CT scan of this motion phantom. For each set of contours, an IMAT plan to be delivered with constant dose rate was created. The plans were generated on a CT scan of the phantom in the static condition with 3 mm PTV margin and applied to the motion phantom under four conditions: static, full superior-inferior (SI) motion (A = 1 cm, T = 4 s) and gating conditions (25% and 50% duty cycles) with full SI motion. A 6 by 15 cm piece of radiographic film was placed in the sagittal plane of the phantom and then irradiated under all measurement conditions. Film calibration was performed with a step-wedge method to convert optical density to dose. Gated IMAT delivery was first validated in 2D by comparing static film with that from gating and full motion. A previously verified simulation tool for IMRT that takes the log files from the multileaf collimator (MLC) controller and the gating system were adapted to simulate the delivered IMAT treatment for full 3D dosimetric analysis. The IMAT simulations were validated against the 2D film measurements. The resultant IMAT simulations were evaluated with dose criteria, dose-volume histograms and 3D gamma analysis. We validated gated IMAT deliveries when we compared the static film with the one from gating using 25% duty cycle using 2D gamma analysis. Within experimental and setup uncertainties, film measurements agreed with their corresponding simulated plans using 2D gamma analysis. Finally, when planning with margins designed for gating with 25% duty cycle and applying 50% or no gating during treatment, the dose differences in D(min,) D(99%) and D(95%) of the clinical target volume can be up to 27 cGy, 20 cGy and 18 cGy, respectively, for a plan with 200 cGy prescription dose. We have experimentally delivered gated IMAT with constant dose rate to a motion phantom and assessed their accuracies with film dosimetry and Monte Carlo simulations. Film dosimetry demonstrated that 25% gating and static plans are within 2%, 2 mm. The Monte Carlo simulation method was employed to generate dose delivered in 3D to a motion phantom, and the dosimetric results were reported. Since our film measurements agreed well with Monte Carlo simulations, we can reliably use this simulation tool to further study the dosimetric effects of target motion and effectiveness of gating for IMAT deliveries.

Incorporating geometric ray tracing to generate initial conditions for intensity modulated arc therapy optimization.

PURPOSE AND BACKGROUNDnIntensity modulated arc therapy (IMAT) is a rotational variant of Intensity modulated radiation therapy (IMRT) that is achieved by allowing the multileaf collimator (MLC) positions to vary as the gantry rotates around the patient. This work describes a method to generate an IMAT plan through the use of a fast ray tracing technique based on dosimetric and geometric information for setting initial MLC leaf positions prior to final IMAT optimization.nnnMETHODS AND MATERIALSnThree steps were used to generate an IMAT plan. The first step was to generate arcs based on anatomical contours. The second step was to generate ray importance factor (RIF) maps by ray tracing the dose distribution inside the planning target volume (PTV) to modify the MLC leaf positions of the anatomical arcs to reduce the maximum dose inside the PTV. The RIF maps were also segmented to create a new set of arcs to improve the dose to low dose voxels within the PTV. In the third step, the MLC leaf positions from all arcs were put through a leaf position optimization (LPO) algorithm and brought into a fast Monte Carlo dose calculation engine for a final dose calculation. The method was applied to two phantom cases, a clinical prostate case and the Radiological Physics Center (RPC)s head and neck phantom. The authors assessed the plan improvements achieved by each step and compared plans with and without using RIF. They also compared the IMAT plan with an IMRT plan for the RPC phantom.nnnRESULTSnAll plans that incorporated RIF and LPO had lower objective function values than those that incorporated LPO only. The objective function value was reduced by about 15% after the generation of RIF arcs and 52% after generation of RIF arcs and leaf position optimization. The IMAT plan for the RPC phantom had similar dose coverage for PTV1 and PTV2 (the same dose volume histogram curves), however, slightly lower dose to the normal tissues compared to a six-field IMRT plan.nnnCONCLUSIONnThe use of a ray importance factor can generate initial IMAT arcs efficiently for further MLC leaf position optimization to obtain more favorable IMAT plan.

Evaluation of optimization strategies and the effect of initial conditions on IMAT optimization using a leaf position optimization algorithm

Intensity-modulated arc therapy (IMAT) is a rotational variant of intensity-modulated radiation therapy (IMRT) that can be implemented with or without angular dose rate variation. The purpose of this study is to assess optimization strategies and initial conditions using a leaf position optimization (LPO) algorithm altered for variable dose rate IMAT. A concave planning target volume (PTV) with a central cylindrical organ at risk (OAR) was used in this study. The initial IMAT arcs were approximated by multiple static beams at 5 degrees angular increments where multi-leaf collimator (MLC) leaf positions were determined from the beams eye view to irradiate the PTV but avoid the OAR. For the optimization strategy, two arcs with arc ranges of 280 degrees and 150 degrees were employed and plans were created using LPO alone, variable dose rate optimization (VDRO) alone, simultaneous LPO and VDRO and sequential combinations of these strategies. To assess the MLC initialization effect, three single 360 degrees arc plans with different initial MLC configurations were generated using the simultaneous LPO and VDRO. The effect of changing optimization degrees of freedom was investigated by employing 3 degrees , 5 degrees and 10 degrees angular sampling intervals for the two 280 degrees , two 150 degrees and single arc plans using LPO and VDRO. The objective function value, a conformity index, a dose homogeneity index, mean dose to OAR and normal tissues were computed and used to evaluate the treatment plans. This study shows that the best optimization strategy for a concave target is to use simultaneous MLC LPO and VDRO. We found that the optimization result is sensitive to the choice of initial MLC aperture shapes suggesting that an LPO-based IMAT plan may not be able to overcome local minima for this geometry. In conclusion, simultaneous MLC leaf position and VDRO are needed with the most appropriate initial conditions (MLC positions, arc ranges and number of arcs) for IMAT.

TU‐E‐ValB‐07: A Segmentation and Leaf Sequencing Algorithm for IMAT

Purpose: To develop an intensity segmentation and leaf sequencing algorithm specifically for intensity‐modulated arc therapy (IMAT), which can be applied to optimized intensity patterns derived from existing commercial IMRT inverse planning software. Methods and Materials: Three phantom cases, as well as a clinical case were planned using a Hi‐Art II (Tomotherapy Inc, WI.) planning station. The end of planning sinograms were then extracted and inputted into our IMAT conversion algorithm. The number of required arcs, deliverable MLC segments for each arc and the relative intensity weighting of each arc were outputted. The number of arcs (modulation) could be controlled by a user parameter, α. The resulting MLC segments were then fed into a fast monte‐carlo dose calculation algorithm, NXEGS (NumeriX, LLC) to obtain 3D dose distributions. Dose statistics (max, min, mean) and dose volume histograms of relevant structures were calculated and compared against the results generated by the Hi‐Art II system. Results: Each plan was converted in under three minutes on a typical desktop PC, with the arc numbers varying between 4 and 15 360° arcs. Qualitatively, the dose distributions obtained from the IMAT plans were similar to the tomotherapy results, as well as planned doses. Quantitatively, the IMAT plans were slightly degraded, with the average dose to normal structures being 7.5% higher for IMAT vs. tomotherapy. However, the IMAT plans generally met planned values, being 9.1% below for maximum doses to normal structures. The number of arcs and therefore the resulting dose distribution could be varied according to α. Conclusions: IMAT segmentation and leaf sequencing produced deliverable IMAT MLC segments and relative arc weights directly from Hi‐Art II optimized plans. The algorithm was computationally efficient, and produced similar dose distributions. Additional optimization could improve resulting dose distributions further. IMAT back‐up for tomotherapy is another potential application.

Poster — Thurs Eve‐40: The potential of using SPECT ventilation information with IMRT for functional lung avoidance in radiotherapy of non small cell lung cancer

We have investigated the feasibility of using ventilation scans obtained from single photon emission computed tomography (SPECT) in intensity-modulated radiation therapy (IMRT) planning in lung cancer radiotherapy to avoid well functioning lung. We fused SPECT ventilation scans acquired at GE Hawkeye SPECT-CT of ten stage-III lung radiotherapy patients with planning CT in treatment planning system (Pinnacle v8.0, Philips Medical Systems). We automatically segment out 50% and 70% ventilated volumes. For each patient, we generated IMRT plans using nine equally spaced beams with and without avoiding well ventilated volume. They were compared with three beam IMRT plans with beam directions chosen to minimize the mean dose to the ventilated lung volumes, while keeping cord dose below tolerance and dose uniformity in the target. The plans generated using functional lung avoidance information reduces the doses to the functioning lung. With both IMRT avoidance plans, we could not obtain better functional avoidance or lower V-20Gy (volume receiving 20Gy or more) for total lung when the planning target volume (PTV) was surrounded by functional lung volumes. We were able to achieve better ventilated lung avoidance and lower total lung V-20Gy when the PTV is close to, but not surrounded by functioning lung volumes. For patients with the PTV that is far from 50% and 70% functional lung volumes, three-field IMRT spare the ventilated lung as well as nine-field IMRT ventilation avoidance plan, with a lower total lung V20-Gy.

SU‐FF‐T‐296: Inverse Treatment Planning Using Volume Sampling with Monte Carlo Dose Calculations

Purpose: We investigated three methods of random sampling of voxels within regions of interest (ROI), using Monte Carlo dose calculations for inverse treatment planning. We studied their effects on file size, accuracy of dose volume histograms, computation time and accuracy in the objective function and determined the impact of number of simulation histories on the objective function. Method and Materials: A dose distribution, stored as double precision, of a clinical lungcancer plan was calculated using Monte Carlo simulation (NXEGS NumeriX, LLC). Only the dose in the ROI (excluding the external) is required, and is stored as integers. Three equations were tested to determine the number of sampling points within each ROI. The first was to keep the same relative percentage of volume for all ROIs. The second was proportional to the hyperbolic tangent of each ROI volume, while the third was proportional to the cube root of the volume. A least square objective function was calculated on all resulting sampling methods. Results: By saving dose values as integers instead of doubles, a 75% reduction is seen in file size, while keeping accuracy to 0.001%. The objective function computational efficiency improvement is directly proportional to the data storage reduction. A further reduction of 94% and 73% occurred when using the cube root and hyperbolic tan of the volume respectively, while producing a 0.03% and 1.8% difference in objective functions compared to that calculated with full ROI volumes. Conclusion: By sampling ROIs where the number of points is proportional to the hyperbolic tangent of the volume there was a savings of approximately 75% of data stored which directly translated into reduction in objective function computational time, with a 0.03% difference in objective function compared to that with full volume calculations.

SU‐DD‐A1‐05: A Ray Tracing Method to Generate Initial Conditions for IMAT Optimization

Purpose: To investigate the utility of using ray tracing to extract intrinsic information from CT, contour and primary dose data in order to determine initial conditions (number of arcs, arc weights, arc ranges and leaf positions) that can be input into an Intensity Modulated Arc Therapy (IMAT) optimization routine. Methods and Materials: Patient CT and contour data was ray‐traced to determine PTV and PTV‐OAR arcs. An additional arc was determined by the calculation of a ray importance factor (RIF) through ray tracing of the primary dose ray‐tracing of the PTV. All three sets of arcs were then input into a previously described leaf position optimization algorithm. This method was tested on two geometries by ray tracing 27 equi‐spaced beams. The optimized arc deliveries (number of arcs, arc weights, arc ranges and leaf positions) were then input into a fast dose calculation algorithm, NXEGS (NumeriX LLC) for dose calculation and comparison with primary dose as calculated by ray tracing.Results: RIF arc addition reduced the objective function by 20% for geometry 1 and 8% for geometry 2. Leaf position optimization further reduced the objective function by 27% for geometry 1 and 29% for geometry 2. Calculation of dose using NXEGS provides accurate dose distributions for IMAT. Conclusions:Ray tracing can quickly provide information about number of arcs, arc ranges, arc weights and leaf positions with very little user input. Leaf position optimization can improve leaf positions once the initial number of arcs and arc ranges are determined. Together these two steps can produce intensity modulated arcs for further optimization with a more accurate dose calculation algorithm.