Optimization of inverse treatment planning using a fuzzy weight function

A fast optimization approach for treatment planning of volumetric modulated arc therapy

Background

Volumetric modulated arc therapy (VMAT) is widely used in clinical practice. It not only significantly reduces treatment time, but also produces high-quality treatment plans. Current optimization approaches heavily rely on stochastic algorithms which are time-consuming and less repeatable. In this study, a novel approach is proposed to provide a high-efficient optimization algorithm for VMAT treatment planning.

Methods

A progressive sampling strategy is employed for beam arrangement of VMAT planning. The initial beams with equal-space are added to the plan in a coarse sampling resolution. Fluence-map optimization and leaf-sequencing are performed for these beams. Then, the coefficients of fluence-maps optimization algorithm are adjusted according to the known fluence maps of these beams. In the next round the sampling resolution is doubled and more beams are added. This process continues until the total number of beams arrived. The performance of VMAT optimization algorithm was evaluated using three clinical cases and compared to those of a commercial planning system.

Results

The dosimetric quality of VMAT plans is equal to or better than the corresponding IMRT plans for three clinical cases. The maximum dose to critical organs is reduced considerably for VMAT plans comparing to those of IMRT plans, especially in the head and neck case. The total number of segments and monitor units are reduced for VMAT plans. For three clinical cases, VMAT optimization takes < 5 min accomplished using proposed approach and is 3–4 times less than that of the commercial system.

Conclusions

The proposed VMAT optimization algorithm is able to produce high-quality VMAT plans efficiently and consistently. It presents a new way to accelerate current optimization process of VMAT planning.

Application programming in C# environment with recorded user software interactions and its application in autopilot of VMAT/IMRT treatment planning

An autopilot scheme of volumetric‐modulated arc therapy (VMAT)/intensity‐modulated radiation therapy (IMRT) planning with the guidance of prior knowledge is established with recorded interactions between a planner and a commercial treatment planning system (TPS). Microsoft (MS) Visual Studio Coded UI is applied to record some common planner‐TPS interactions as subroutines. The TPS used in this study is a Windows‐based Eclipse system. The interactions of our application program with Eclipse TPS are realized through a series of subroutines obtained by prerecording the mouse clicks or keyboard strokes of a planner in operating the TPS. A strategy to autopilot Eclipse VMAT/IMRT plan selection process is developed as a specific example of the proposed “scripting” method. The autopiloted planning is navigated by a decision function constructed with a reference plan that has the same prescription and similar anatomy with the case at hand. The calculation proceeds by alternating between the Eclipse optimization and the outer‐loop optimization independent of the Eclipse. In the C# program, the dosimetric characteristics of a reference treatment plan are used to assess and modify the Eclipse planning parameters and to guide the search for a clinically sensible treatment plan. The approach is applied to plan a head and neck (HN) VMAT case and a prostate IMRT case. Our study demonstrated the feasibility of application programming method in C# environment with recorded interactions of planner‐TPS. The process mimics a planner's planning process and automatically provides clinically sensible treatment plans that would otherwise require a large amount of manual trial and error of a planner. The proposed technique enables us to harness a commercial TPS by application programming via the use of recorded human computer interactions and provides an effective tool to greatly facilitate the treatment planning process.

PACS number(s): 87.55.D‐, 87.55.kd, 87.55.de

Intelligence-guided beam angle optimization in treatment planning of intensity-modulated radiation therapy

An intelligence guided approach based on fuzzy inference system (FIS) was proposed to automate beam angle optimization in treatment planning of intensity-modulated radiation therapy (IMRT). The model of FIS is built on inference rules in describing the relationship between dose quality of IMRT plan and irradiated region of anatomical structure. Dose quality of IMRT plan is quantified by the difference between calculated and constraint doses of the anatomical structures in an IMRT plan. Irradiated region of anatomical structure is characterized by the metric, covered region of interest, which is the region of an anatomical structure under radiation field while beam's eye-view is conform to target volume. Initially, an IMRT plan is created with a single beam. The dose difference is calculated for the input of FIS and the output of FIS is obtained with processing of fuzzy inference. Later, a set of candidate beams is generated for replacing the current beam. This process continues until no candidate beams is found. Then the next beam is added to the IMRT plan and optimized in the same way as the previous beam. The new beam keeps adding to the IMRT plan until the allowed beam number is reached. Two spinal cases were investigated in this study. The preliminary results show that dose quality of IMRT plans achieved by this approach is better than those achieved by the default approach with equally spaced beam setting. It is effective to find the optimal beam combination of IMRT plan with the intelligence-guided approach.

Development of a neuro-fuzzy technique for automated parameter optimization of inverse treatment planning

Background

Parameter optimization in the process of inverse treatment planning for intensity modulated radiation therapy (IMRT) is mainly conducted by human planners in order to create a plan with the desired dose distribution. To automate this tedious process, an artificial intelligence (AI) guided system was developed and examined.

Methods

The AI system can automatically accomplish the optimization process based on prior knowledge operated by several fuzzy inference systems (FIS). Prior knowledge, which was collected from human planners during their routine trial-and-error process of inverse planning, has first to be "translated" to a set of "if-then rules" for driving the FISs. To minimize subjective error which could be costly during this knowledge acquisition process, it is necessary to find a quantitative method to automatically accomplish this task. A well-developed machine learning technique, based on an adaptive neuro fuzzy inference system (ANFIS), was introduced in this study. Based on this approach, prior knowledge of a fuzzy inference system can be quickly collected from observation data (clinically used constraints). The learning capability and the accuracy of such a system were analyzed by generating multiple FIS from data collected from an AI system with known settings and rules.

Results

Multiple analyses showed good agreements of FIS and ANFIS according to rules (error of the output values of ANFIS based on the training data from FIS of 7.77 ± 0.02%) and membership functions (3.9%), thus suggesting that the "behavior" of an FIS can be propagated to another, based on this process. The initial experimental results on a clinical case showed that ANFIS is an effective way to build FIS from practical data, and analysis of ANFIS and FIS with clinical cases showed good planning results provided by ANFIS. OAR volumes encompassed by characteristic percentages of isodoses were reduced by a mean of between 0 and 28%.

Conclusion

The study demonstrated a feasible way to automatically perform parameter optimization of inverse treatment planning under guidance of prior knowledge without human intervention other than providing a set of constraints that have proven clinically useful in a given setting.

The application of distance transformation on parameter optimization of inverse planning in intensity-modulated radiation therapy

In inverse planning for intensity‐modulated radiation therapy (IMRT), the dose specification and related weighting factor of an objective function for involved organs is usually predefined by a single value and then iteratively optimized, subject to a set of dose—volume constraints. Because the actual dose distribution is essentially non‐uniform and considerably affected by the geometric shape and distribution of the anatomic structures involved, the spatial information regarding those structures should be incorporated such that the predefined parameter distribution is made to approach the clinically expected distribution. Ideally, these parameter distributions should be predefined on a voxel basis in a manual method. However, such an approach is too time‐consuming to be feasible in routine use.

In the present study, we developed a computer‐aided method to achieve the goal described above, producing a non‐uniform parameter distribution based on spatial information about the anatomic structures involved. The method consists of two steps:

Use distance transformation technique to calculate the distance distribution of the structures.

Based on the distance distribution, produce the parameter distribution via a conversion function guided by prior knowledge.

We use two simulated cases to examine the effectiveness of the method. The results indicate that application of a non‐uniform parameter distribution produced by distance transformation clearly improves dose‐sparing of critical organs without compromising dose coverage of the planning target.

PACS numbers: 87.53.Jw

Evaluation of an artificial intelligence guided inverse planning system: Clinical case study

PURPOSE

An artificial intelligence (AI) guided method for parameter adjustment of inverse planning was implemented on a commercial inverse treatment planning system. For evaluation purpose, four typical clinical cases were tested and the results from both plans achieved by automated and manual methods were compared.

METHODS AND MATERIALS

The procedure of parameter adjustment mainly consists of three major loops. Each loop is in charge of modifying parameters of one category, which is carried out by a specially customized fuzzy inference system. A physician prescribed multiple constraints for a selected volume were adopted to account for the tradeoff between prescription dose to the PTV and dose-volume constraints for critical organs. The searching process for an optimal parameter combination began with the first constraint, and proceeds to the next until a plan with acceptable dose was achieved. The initial setup of the plan parameters was the same for each case and was adjusted independently by both manual and automated methods. After the parameters of one category were updated, the intensity maps of all fields were re-optimized and the plan dose was subsequently re-calculated. When final plan arrived, the dose statistics were calculated from both plans and compared.

RESULTS

For planned target volume (PTV), the dose for 95% volume is up to 10% higher in plans using the automated method than those using the manual method. For critical organs, an average decrease of the plan dose was achieved. However, the automated method cannot improve the plan dose for some critical organs due to limitations of the inference rules currently employed. For normal tissue, there was no significant difference between plan doses achieved by either automated or manual method.

CONCLUSION

With the application of AI-guided method, the basic parameter adjustment task can be accomplished automatically and a comparable plan dose was achieved in comparison with that achieved by the manual method. Future improvements to incorporate case-specific inference rules are essential to fully automate the inverse planning process.

Inverse planning in three-dimensional conformal and intensity-modulated radiotherapy of mid-thoracic oesophageal cancer

The aim of this study is to demonstrate the use of inverse planning in three-dimensional conformal radiation therapy (3DCRT) of oesophageal cancer patients and to evaluate its dosimetric results by comparing them with forward planning of 3DCRT and inverse planning of intensity-modulated radiotherapy (IMRT). For each of the 15 oesophageal cancer patients in this study, the forward 3DCRT, inverse 3DCRT and inverse IMRT plans were produced using the FOCUS treatment planning system. The dosimetric results and the planner's time associated with each of the treatment plans were recorded for comparison. The inverse 3DCRT plans showed similar dosimetric results to the forward plans in the planning target volume (PTV) and organs at risk (OARs). However, they were inferior to that of the IMRT plans in terms of tumour control probability and target dose conformity. Furthermore, the inverse 3DCRT plans were less effective in reducing the percentage lung volume receiving a dose below 25 Gy when compared with the IMRT plans. The inverse 3DCRT plans delivered a similar heart dose as in the forward plans, but higher dose than the IMRT plans. The inverse 3DCRT plans significantly reduced the operator's time by 2.5 fold relative to the forward plans. In conclusion, inverse planning for 3DCRT is a reasonable alternative to the forward planning for oesophageal cancer patients with reduction of the operator's time. However, IMRT has the better potential to allow further dose escalation and improvement of tumour control.

Fuzzy logic guided inverse treatment planning

A fuzzy logic technique was applied to optimize the weighting factors in the objective function of an inverse treatment planning system for intensity-modulated radiation therapy (IMRT). Based on this technique, the optimization of weighting factors is guided by the fuzzy rules while the intensity spectrum is optimized by a fast-monotonic-descent method. The resultant fuzzy logic guided inverse planning system is capable of finding the optimal combination of weighting factors for different anatomical structures involved in treatment planning. This system was tested using one simulated (but clinically relevant) case and one clinical case. The results indicate that the optimal balance between the target dose and the critical organ dose is achieved by a refined combination of weighting factors. With the help of fuzzy inference, the efficiency and effectiveness of inverse planning for IMRT are substantially improved.

Multiobjective inverse planning for intensity modulated radiotherapy with constraint-free gradient-based optimization algorithms

We consider the behaviour of the limited memory L-BFGS algorithm as a representative constraint-free gradient-based algorithm which is used for multiobjective (MO) dose optimization for intensity modulated radiotherapy (IMRT). Using a parameter transformation, the positivity constraint problem of negative beam fluences is entirely eliminated: a feature which to date has not been fully understood by all investigators. We analyse the global convergence properties of L-BFGS by searching for the existence and the influence of possible local minima. With a fast simulated annealing (FSA) algorithm we examine whether the L-BFGS solutions are globally Pareto optimal. The three examples used in our analysis are a brain tumour, a prostate tumour and a test case with a C-shaped PTV. In 1% of the optimizations global convergence is violated. A simple mechanism practically eliminates the influence of this failure and the obtained solutions are globally optimal. A single-objective dose optimization requires less than 4 s for 5400 parameters and 40000 sampling points. The elimination of the problem of negative beam fluences and the high computational speed permit constraint-free gradient-based optimization algorithms to be used for MO dose optimization. In this situation, a representative spectrum of possible solutions is obtained which contains information such as the trade-off between the objectives and range of dose values. Using simple decision making tools the best of all the possible solutions can be chosen. We perform an MO dose optimization for the three examples and compare the spectra of solutions, firstly using recommended critical dose values for the organs at risk and secondly, setting these dose values to zero.

AI-guided parameter optimization in inverse treatment planning

An artificial intelligence (AI)-guided inverse planning system was developed to optimize the combination of parameters in the objective function for intensity-modulated radiation therapy (IMRT). In this system, the empirical knowledge of inverse planning was formulated with fuzzy if-then rules, which then guide the parameter modification based on the on-line calculated dose. Three kinds of parameters (weighting factor, dose specification, and dose prescription) were automatically modified using the fuzzy inference system (FIS). The performance of the AI-guided inverse planning system (AIGIPS) was examined using the simulated and clinical examples. Preliminary results indicate that the expected dose distribution was automatically achieved using the AI-guided inverse planning system, with the complicated compromising between different parameters accomplished by the fuzzy inference technique. The AIGIPS provides a highly promising method to replace the current trial-and-error approach.

Beam orientation optimization for IMRT by a hybrid method of the genetic algorithm and the simulated dynamics

We have developed a new method for beam orientation optimization in intensity-modulated radiation therapy (IMRT). The problem of beam orientation optimization in IMRT is solved by a decoupled two-step iterative process: (1) optimization of the intensity profiles for given beam configurations; (2) selection of optimal beam configurations based on the ranking by an objective function score for the results of the intensity profile optimization. The simulated dynamics algorithm is used for the intensity profile optimization. This algorithm enforces both the hard constraints and dose-volume constraints. A genetic algorithm is used to select beam orientation configurations. The method has been tested for both a simulated and clinical case, and the results show that beam orientation optimization significantly improved IMRT plans within a time period that is clinically acceptable. The results also show the dependence of the optimal orientation configurations on the prescribed constraints. In addition, beam orientation optimization by the method described here can provide multiple plans with similar dose distributions. This degeneracy characteristic can be exploited to our advantage in introducing additional planning objectives, e.g., the smoothness of intensity profiles, for the selection of the optimal plan among the degenerate configurations for treatment delivery.

Therapeutic treatment plan optimization with probability density-based dose prescription

The dose optimization in inverse planning is realized under the guidance of an objective function. The prescription doses in a conventional approach are usually rigid values, defining in most instances an ill-conditioned optimization problem. In this work, we propose a more general dose optimization scheme based on a statistical formalism [Xing et al., Med. Phys. 21, 2348-2358 (1999)]. Instead of a rigid dose, the prescription to a structure is specified by a preference function, which describes the user's preference over other doses in case the most desired dose is not attainable. The variation range of the prescription dose and the shape of the preference function are predesigned by the user based on prior clinical experience. Consequently, during the iterative optimization process, the prescription dose is allowed to deviate, with a certain preference level, from the most desired dose. By not restricting the prescription dose to a fixed value, the optimization problem becomes less ill-defined. The conventional inverse planning algorithm represents a special case of the new formalism. An iterative dose optimization algorithm is used to optimize the system. The performance of the proposed technique is systematically studied using a hypothetical C-shaped tumor with an abutting circular critical structure and a prostate case. It is shown that the final dose distribution can be manipulated flexibly by tuning the shape of the preference function and that using a preference function can lead to optimized dose distributions in accordance with the planner's specification. The proposed framework offers an effective mechanism to formalize the planner's priorities over different possible clinical scenarios and incorporate them into dose optimization. The enhanced control over the final plan may greatly facilitate the IMRT treatment planning process.

An optimization algorithm for intensity modulated radiotherapy--the simulated dynamics with dose-volume constraints

We have developed a new method for optimization in intensity modulated radiation therapy (IMRT) that makes use of simulated dynamics in a classical system of interacting particles. An analogy is drawn between intensity profile optimization in IMRT and relaxation to the equilibrium configuration in a dynamic system. The intensities of beamlets are equivalent to the positions of the virtual particles. The potential energy of the system is defined by the objective function, which determines the equations of motion for the virtual particles. In this paper, we present the implementation of dose constraints and dose-volume constraints. Our strategy is to optimize the dose to the planned target volume (PTV) while keeping all constraints to the organs at risk (OARs) satisfied rigorously. A simple quadratic objective function is used that only includes terms for PTV voxels. By this approach, no additional parameters other than that for prescribing desired dose and constraints, such as importance factors, are needed. The hard constraints that require non-negative beamlet intensities and that the dose at any voxel in an OAR cannot exceed a maximum tolerance, are implemented as semi-transmittable potential barriers of infinite height. Dose-volume constraints are handled by placing hard constraints on partial volumes. Handling of the clinically applied constraints was tested using phantoms and clinical cases. Our results show that the dose-volume histogram (DVH) type of plan prescription can be fulfilled with satisfactory PTV coverage. In addition to the convenience of implementation, our method can achieve a high computational efficiency with the understanding of the dynamic behavior of the system.

Inverse planning for functional image-guided intensity-modulated radiation therapy

Radiation therapy is an image-guided process whose success critically depends on the imaging modality used for treatment planning and the level of integration of the available imaging information. In this work, we establish a dose optimization framework for incorporating metabolic information from functional imaging modalities into the intensity-modulated radiation therapy (IMRT) inverse planning process and to demonstrate the technical feasibility of planning deliberately non-uniform dose distributions in accordance with functional imaging data. For this purpose, a metabolic map from functional images is discretized into a number of abnormality levels (ALs) and then fused with CT images. To escalate dose to the metabolically abnormal regions, we assume, for a given spatial point, a linear relation between the AL and the prescribed dose. But the formalism developed here is independent of the assumption and any other relation between AL and prescription is applicable. For a given AL and prescription relation, it is only necessary to prescribe the dose to the lowest AL in the target and the desired doses to other regions with higher AL values are scaled accordingly. To accomplish differential sparing of a sensitive structure when its functional importance (FI) distribution is known, we individualize the tolerance doses of the voxels within the structure according to their Fl levels. An iterative inverse planning algorithm in voxel domain is used to optimize the system with in homogeneous dose prescription. To model intra-structural trade-off, a mechanism is introduced through the use of voxel-dependent weighting factors, in addition to the conventional structure specific weighting factors which model the inter-structural trade-off. The system is used to plan a phantom case with a few hypothetical functional distributions and a brain tumour treatment with incorporation of magnetic resonance spectroscopic imaging data. The results indicated that it is technically feasible to produce deliberately non-uniform dose distributions according to the functional imaging requirements. Integration of functional imaging information into radiation therapy dose optimization allows for consideration of patient-specific biologic information and provides a significant opportunity to truly individualize radiation treatment. This should enhance our capability to safely and intelligently escalate dose and lays the technical foundation for future clinical studies of the efficacy of functional imaging-guided IMRT.

Molecular dynamics used in radiation therapy

In this Letter, classical molecular dynamics is used to deal with optimization problems occurring in intensity modulated radiation therapy. By introducing the concepts of virtual atom and virtual cluster, the optimization process in this kind of therapy can be considered as an analogy to finding the equilibrium configuration of a cluster. This viewpoint gives great insight into the optimization problems. To show how the idea works, a dose-based objective function is adopted to obtain the optimized intensity profiles. The results and high computational efficiency show that molecular dynamics is applicable clinically for this therapy. The idea presented here also could be inspiring to other fields where optimization problems exist.