Rapid dose calculations for stereotactic radiosurgery

Linear programming based on neural networks for radiotherapy treatment planning

In this paper, we propose a neural network model for linear programming that is designed to optimize radiotherapy treatment planning (RTP). This kind of neural network can be easily implemented by using a kind of 'neural' electronic system in order to obtain an optimization solution in real time. We first give an introduction to the RTP problem and construct a non-constraint objective function for the neural network model. We adopt a gradient algorithm to minimize the objective function and design the structure of the neural network for RTP. Compared to traditional linear programming methods, this neural network model can reduce the time needed for convergence, the size of problems (i.e., the number of variables to be searched) and the number of extra slack and surplus variables needed. We obtained a set of optimized beam weights that result in a better dose distribution as compared to that obtained using the simplex algorithm under the same initial condition. The example presented in this paper shows that this model is feasible in three-dimensional RTP.

An efficient method for small field treatment dose calculation for stereotactic radiosurgery using a LINAC

The normal procedure for a physician-physicist team designing a treatment plan for multiarc stereotactic radiosurgery is the trial-and-error approach of changing the collimator size and the location of the isocenter of radiation and viewing the isodose curves on two-dimensional computed tomography (CT) or magnetic resonance imaging (MRI) image planes. Automatic optimization procedures have also been used to optimize beam weight or beam size. However, either process is very time consuming. To improve the speed of the dose calculation, a random sampling method has been proposed. Unfortunately, the sampled values of an objective function are different from one sample to another. Such a sampling method cannot be used in automatic optimization because the next move in an optimization process is based on the current and past objective function values. To this end, an adaptive method based on the size of the collimators is proposed and used to determine a small volume in the shape of a hollow sphere for which the dose is calculated. With an appropriate choice of an adaptive hollow sphere, the objective function calculated based on such a hollow sphere is the same as that calculated with the traditional three-dimensional (3-D) cube matrix. However, with the new adaptive method, the speed of calculating a dose can be improved by a factor of 4 to a factor of 100. Because of the improvement in the speed of calculating a treatment dose, the new adaptive hollow sphere method for calculating a treatment dose can be used routinely in designing a treatment plan.

A fast algorithm for the dose calculation of irregularly shaped fields in stereotactic convergent beam irradiation

A fast, irregular field dose calculation algorithm is described for use in stereotactic convergent beam irradiation (CBI) with high-energy photons. The dose distribution of an irregular field is approximated by a convolution of narrow pencil beams. Decomposition of pencil beams and a 'pre-selection' algorithm is used to avoid redundant calculations. A typical dose calculation for CBI with nine couch angles and nine different irregularly shaped fields takes less than 90 s on a DEC VAXstation 3100. Fortran. The difference between the measured and the calculated dose values increases with increasing distance from the isocentre. The error in 10% isodose line is less than 3% of the isocentre dose.