Dynamic fluence map sequencing using piecewise linear leaf position functions

Matthew Kelly,Jacobus H.M. van Amerongen,Marleen Balvert,David Craft

DOI: https://doi.org/10.48550/arXiv.1901.03626

2019-01-12

Abstract:Within the setting of intensity modulated radiation therapy (IMRT) and the fully continuous version of IMRT called volumetric modulated radiation therapy (VMAT), we consider the problem of matching a given fluence map as well as possible in limited time by the use of a linear accelerator (linac) with a multi-leaf collimator (MLC). We introduce two modeling strategies to manage the nonconvexity and the associated local minima of this problem. The first is the use of linear splines to model the MLC leaf positions as functions of time. The second is a progressively controllable smooth model (instead of a step function) of how the leaves block the fluence radiation. We propose a two part solution: an outer loop that optimizes the dose rate pattern over time, and an inner loop that given a dose rate pattern optimizes the leaf trajectories.

Medical Physics

What problem does this paper attempt to address?

### Problems the paper attempts to solve

The paper aims to solve a key problem in intensity - modulated radiation therapy (IMRT) and volumetric - modulated arc therapy (VMAT): how to match a given fluence map as precisely as possible using a linear accelerator (Linac) and a multi - leaf collimator (MLC) within a limited time. Specifically, the paper focuses on how to optimize the position trajectory of MLC leaves and the dose - rate pattern to maximize the matching accuracy of the fluence map under the premise of meeting machine - limited conditions.

### Background and motivation

1. **Importance of time**:
- Shorter treatment time can improve the efficiency of treatment facilities, especially in developing countries where these devices are very expensive.
- The longer the patient is under radiation, the greater the impact of movement on the dose distribution.
- There is a trade - off between dose quality and treatment time, so an optimal solution needs to be found.

2. **Limitations of existing methods**:
- **Sliding window least - squares (SWLS)**: Although it can perfectly replicate the fluence map in sufficient time, it is usually not very efficient.
- **Step - and - shoot method**: It is not suitable for continuous treatment and cannot guarantee an optimal solution.

### Research methods

1. **Modeling strategies**:
- **Linear spline function**: Used to represent the change of MLC leaf position over time, thereby managing non - convexity and local minimum problems.
- **Smooth occlusion function**: Replaces the traditional step function to represent how the leaves block radiation in a smoother way.

2. **Optimization framework**:
- **Two - stage optimization**:
- **Outer loop**: Optimizes the change of the dose - rate pattern over time.
- **Inner loop**: Optimizes the leaf trajectory given the dose - rate pattern.

3. **Specific implementation**:
- **Spline representation**: Represents the leaf position using a linear spline function, which is defined by node values.
- **Integral calculation**: Calculates the integral in the objective function using the smooth occlusion function and standard numerical integration methods (such as the midpoint rule).
- **Non - linear programming**: Transforms the problem into a standard non - linear programming problem and solves it using solvers such as FMINCON, SNOPT or IPOPT.

### Experimental results

1. **Optimization of single - row fluence maps**:
- Experiments were carried out using a prostate patient data set to show the optimization results of single - row fluence maps.
- Optimizations were carried out for unimodal and bimodal fluence maps respectively, showing the performance under different smoothing parameters.

2. **Matching of the entire fluence map**:
- By parallelly optimizing the trajectory of each pair of leaves, fluence map matching results at different treatment times were generated.
- The results show that near - perfect fluence map matching can be achieved in a relatively short treatment time (such as 5 seconds).

### Conclusions

- **Main contributions**:
- Proposed a new method using linear spline functions and smooth occlusion functions to optimize MLC leaf trajectories and dose - rate patterns.
- Improved the convergence and accuracy of optimization by iteratively refining the smoothing parameters.
- In clinical practical applications, high - precision fluence map matching can be achieved in a relatively short treatment time.

- **Future work**:
- Further study the dose - optimization problem, combined with dynamic fluence - map sequencing to achieve a more comprehensive optimization scheme.
- Explore more practical application scenarios to verify the clinical feasibility of the method.