GEMSS RAPT Project Details Consortium Reports & Presentations Software Collaboration

Radiosurgery and the Monte Carlo Method (RAPT)

Radiosurgery and GammaPlan

Lars Leksell, a Swedish professor of neurosurgery, introduced stereotactic radiosurgery in the 1950's. He developed a dedicated treatment unit (the Gamma Knife) that uses the cross-firing beams from 201 Cobalt-60 sources to simultaneously irradiate an intracranial lesion [1]. Worldwide, the technique has expanded exponentially with over 190 Gamma Knives currently in operation. These units treated a total of 250,000 patients in the year 2003, compared with 7,000 in 1991. Use of the Gamma Knife therapy unit is assisted by GammaPlan software, which is used for predicting radiation doses delivered to the patient by the Gamma Knife. GammaPlan runs on a Unix workstation and assists an expert operator in determining the sequence of radiation delivery necessary to treat the lesion. The GammaPlan software supplied with the Gamma Knife is fast and effective through its use of perceptive assumptions that describe attenuation of photon beams within the head. However, there are conditions in which the assumptions are invalid (e.g. tissue inhomogeneity) and in these circumstances the GammaPlan simulation provides a flawed description of patient dose.


Dose contours - GammaPlan

The Role of Monte Carlo (RAPT)

A more rigorous description of the radiation distribution can be obtained using complex, compute-intense Monte Carlo simulations [2,3] but this requires a large computational resource. The improved physics of the Monte Carlo method is particularly apparent at boundaries between materials with different linear attenuation coefficients within the head and may indicate significant deviation from solutions obtained using GammaPlan (particularly relevant if a high dose region is close to sensitive structures such as the eyes). This is the motivation for Monte Carlo-based planning as implemented in GEMSS, and custom software developed for the purpose is used (i.e. RAPT - Radiotherapy Application on Parallel Technology - developed by IT Innovation, Southampton, UK). Ultimately, improved confidence in the dose distribution will precipitate more effective plans and treatment, with concomitant improvement in patient management and patient outcome. However, the complex physics modelled by the Monte Carlo method requires long solution times and computation is insufficiently responsive on conventional platforms to provide rapid dose distributions for the planning process - hence the need for the GEMSS Grid.

Monte Carlo

Dose contours - RAPT


  1. L. Leksell, The stereotactic method and radiosurgery of the brain, Acta Chirurgica Scandinavica 102 (1951), 316-319
  2. D.W.O. Rogers, Monte Carlo techniques in radiotherapy, Physics In Canada, Medical Physics Special Issue 58(2) 2002, 63-70
  3. A. Wu, G. Linder, A.H. Maitz, A.M. Kalend, L.D. Lunsford, J.C. Flickinger, W.D. Bloomer, Physics of Gamma Knife approach on convergent beams in stereotactic radiosurgery, International Journal of Radiation Oncology Biology Physics 18(4) 1990, 941-949
[an error occurred while processing this directive]