Errors in inverse treatment planning based on inaccurate dose calculation

Abstract

Inverse treatment planning, which enables intensity modulated radiation treatment (EV1RT), is quickly entering into clinical practice. One of the key ingredients of any inverse, as well as conventional treatment planning systems, is the used dose calculation algorithm. Ideally, the dose calculation algorithm should be accurate and fast; however, there is always some trade off between accuracy and speed. Most of the current inverse treatment algorithms use fast dose calculation algorithms, usually based on pencil beam or convolution/ superposition approach [1–4]. These algorithms work well for many clinical cases, but there are some examples where systematic errors of several percent may occur [5,6]. On the other hand, Monte Carlo based dose calculations are known for their accuracy, but they are also burdened with some statistical error, which introduces different type of problems [7].