Plan averaging for multicriteria navigation of sliding window IMRT and VMAT

Craft, David; Papp, Dávid; Unkelbach, Jan

doi:10.1118/1.4859295

Cited by 14 publications

(14 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In Ref. 44, it is shown that sliding window type delivery allows for VMAT plan averaging, which represents a potential advantage for MCO planning.…”

Section: D Multicriteria Capabilitiesmentioning

confidence: 99%

Optimization approaches to volumetric modulated arc therapy planning

et al. 2015

Self Cite

View full text Add to dashboard Cite

Volumetric modulated arc therapy (VMAT) has found widespread clinical application in recent years. A large number of treatment planning studies have evaluated the potential for VMAT for different disease sites based on the currently available commercial implementations of VMAT planning. In contrast, literature on the underlying mathematical optimization methods used in treatment planning is scarce. VMAT planning represents a challenging large scale optimization problem. In contrast to fluence map optimization in intensity-modulated radiotherapy planning for static beams, VMAT planning represents a nonconvex optimization problem. In this paper, the authors review the state-of-the-art in VMAT planning from an algorithmic perspective. Different approaches to VMAT optimization, including arc sequencing methods, extensions of direct aperture optimization, and direct optimization of leaf trajectories are reviewed. Their advantages and limitations are outlined and recommendations for improvements are discussed. C 2015 American Association of Physicists in Medicine. [http://dx

show abstract

“…In Ref. 44, it is shown that sliding window type delivery allows for VMAT plan averaging, which represents a potential advantage for MCO planning.…”

Section: D Multicriteria Capabilitiesmentioning

confidence: 99%

Optimization approaches to volumetric modulated arc therapy planning

et al. 2015

Self Cite

View full text Add to dashboard Cite

show abstract

“…The resulting expressions from the application of Eqs. (1), (2) and (3) to the different fits describing the respective dose distributions are summarized in Table I for different PDFs. A graphical representation of the fitted dose profile and a rectangular as well as a Gaussian PDF are shown in Fig.…”

Section: A General Formalismmentioning

confidence: 99%

“…To investigate the propagation of uncertainties, various convolutions of rectangular and Gaussian PDFs were performed and the resulting PDFs were applied to the formalism Medical Physics, 47 (3), March 2020 described in Eqs. (1) and (2). The resulting equations for the one dimensional case are summarized in Tables II and III.…”

Section: B Uncertainty Propagationmentioning

confidence: 99%

“…Depending on the size and complexity of the target volume on the one hand and the employed treatment technique, as well as the sequencing algorithms on the other hand, the frequency of small fields compared to the total fluence can vary. [1][2][3][4][5] Using small fields in radiation oncology requires a state of the art treatment planning system suitable for this task. For beam model creation and commissioning of such a system, special procedures on the determination of field output factors in the small fields need to be considered.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

An analytical formalism for the assessment of dose uncertainties due to positioning uncertainties

Lechner

Georg

2020

Medical Physics

View full text Add to dashboard Cite

Purpose To present an analytical formalism for the in depth assessment of uncertainties of field output factors in small fields related to detector positioning based on dose profile measurements. Additionally, a procedure for the propagation of these uncertainties was developed. Methods Based on the assumption that one dimensional and two dimensional second‐order polynomial functions can be fitted to dose profiles of small photon beams, equations for the calculation of the expectation value, the variance, and the standard deviation were developed. The following fitting procedures of the dose profiles were considered: A one‐dimensional case (1D), a quasi two‐dimensional case (2Dq) based on independently measured line profiles and a full 2D case (2Df) which also considers cross‐correlations in a two‐dimensional dose distribution. A rectangular and a Gaussian probability density function (PDF) characterizing the probability of possible positions of the detector relative to the maximum dose were used. Uncertainty components such as the finite resolution of the scanning water phantom, the reproducibility of the determination of the position of the maximum dose, and the reproducibility of the collimator system were investigated. This formalism was tested in a 0.5 x 0.5 cm2 photon field where dose profiles were measured using a radiochromic film, a synthetic diamond detector, and an unshielded diode detector. Additionally, the dose distribution measured with the radiochromic film was convoluted with a convolution kernel mimicking the active volume of the unshielded diode. Results Analytic expressions for the calculation of uncertainties on field output factors were found for the 1D, the 2Dq, and the 2Df case. The uncertainty of the field output factor related to the relative position of the detector to the maximum dose increased quadratically with increasing limits of possible detector positions. Analysis of the radiochromic film showed that the 2Dq case gave a more conservative assessment of the uncertainty compared to the 2Df case with a difference of < 0.1%. The 2Dq case applied to the film measurements agreed well with the same approach as was applied to the unshielded diode. The investigated uncertainty components propagated to an uncertainty of the field output factors of 0.5% and 0.4% for the synthetic diamond and the unshielded diode, respectively. Additionally, the expectation value was lower than the maximum dose. The difference was 0.4% and 0.3% for the synthetic diamond and the unshielded diode, respectively. Conclusions The assessment of uncertainties of field output factors related to detector positioning is feasible using the proposed formalism. The 2Dq case is applicable when using online detectors. Accurate positioning in small fields is essential for accurate dosimetry as its related uncertainty increases quadratically. The observed drop of the expectation value needs to be considered in small field dosimetry.

show abstract

“…Moreover, these parameters are subject to strong hardware constraints, which have to be satisfied by a treatment plan to ensure its deliverability. Over the years, a number of research efforts have been devoted to developing novel algorithms to solve the VMAT optimization problem [8][9][10][11][12][13][14][15][16][17][18][19][20][21][22][23][24] . In particular, we have previously developed a column-generation approach 16,18 , and our test results on real patient cases have demonstrated the potential of this method in terms of generating a plan with all the degrees of freedom considered.…”

Section: Introductionmentioning

confidence: 99%

Multi-GPU implementation of a VMAT treatment plan optimization algorithm

et al. 2015

View full text Add to dashboard Cite

Purpose: VMAT optimization is a computationally challenging problem due to its large data size, high degrees of freedom, and many hardware constraints. Highperformance graphics processing units (GPUs) have been used to speed up the computations. However, GPU's relatively small memory size cannot handle cases with a large dose-deposition coefficient (DDC) matrix in cases of, e.g., those with a large target size, multiple targets, multiple arcs and/or small beamlet size. The main purpose of this paper is to report an implementation of a column-generation based VMAT algorithm, previously developed in our group, on a multi-GPU platform to solve the memory limitation problem. While the column-generation based VMAT algorithm has been previously developed, the GPU implementation details have not been reported. Hence, another purpose is to present detailed techniques employed for GPU implementation. We also would like to utilize this particular problem as an example problem to study the feasibility of using a multi-GPU platform to solve large-scale problems in medical physics. Methods: The column-generation approach generates VMAT apertures sequentially by solving a pricing problem (PP) and a master problem (MP) iteratively. In our method, the sparse DDC matrix is first stored on CPU in coordinate list format (COO). On the GPU side, this matrix is split into four sub-matrices according to beam angles, which are stored on four GPUs in compressed sparse row (CSR) format. Computation of beamlet price, the first step in PP, is accomplished using multi-GPUs. A fast inter-GPU data transfer scheme is designed using peer-to-peer (P2P) access. The remaining steps of PP and MP problems are implemented on CPU or a single GPU due to their modest problem scale and computational loads. Barzilai and Borwein (BB) algorithm with subspace step scheme is adopted here to solve the MP problem. A head and neck (H&N) cancer case was used to validate our method. We also compare our multi-GPU implementation with three different single GPU implementation strategies: truncating DDC matrix (S1), repeatedly transferring DDC matrix between CPU and GPU (S2), and porting computations involving DDC matrix to CPU (S3), in terms of both plan quality and computational efficiency. Two 2 more H&N patient cases and three prostate cases were also used to demonstrate the advantages of our method. Results: Our multi-GPU implementation can finish the optimization process within ~1 minute for the H&N patient case. S1 leads to an inferior plan quality although its total time was 10 seconds shorter than the multi-GPU implementation due to the reduced matrix size. S2 and S3 yield the same plan quality as the multi-GPU implementation but take ~4 minutes and ~6 minutes, respectively. High computational efficiency was consistently achieved for the other 5 patient cases tested, with VMAT plans of clinically acceptable quality obtained within 23~46 seconds. Conversely, to obtain clinically comparable or acceptable plans for all these 6 VMAT cases that we have tested in t...

show abstract

Plan averaging for multicriteria navigation of sliding window IMRT and VMAT

Cited by 14 publications

References 22 publications

Optimization approaches to volumetric modulated arc therapy planning

Optimization approaches to volumetric modulated arc therapy planning

An analytical formalism for the assessment of dose uncertainties due to positioning uncertainties

Multi-GPU implementation of a VMAT treatment plan optimization algorithm

Contact Info

Product

Resources

About