Interactive Treatment Planning For Radiosurgery

 

 

Motivation

Radiotherapy is an important tool in cancer treatment and management. The main idea is to damage the cancerous tissue by exposing it to large amounts of radiation. The very nature of this treatment implies that healthy tissue and organs will be subject to significant doses of radiation.

Traditionally, the task of tweaking the optimization constraints is achieved by analyzing the cumulative dose-volume histograms (DVH) for each volume of interest (VOI) and selecting the dose bounds accordingly. While easy to interpret and offering a good overview of the dose coverage of a particular VOI, a DVH offers no spatial information about the dose distribution.

 

Research Direction

We propose a novel approach in which the planner operates in a three-dimensional space by interacting with the iso-dose surface. Different interactions such as pushing or pulling of the surface are possible, as well as explicitly relaxing the dose bounds for the corresponding voxels. This enables the clinician to either specifically cover or spare these areas from being covered by the iso-dose surface.

The prostate (1), rectum (2), bladder (3) and the iso-dose surface at 36 Gy before (left) manipulation, during (center) the use of the pull tool and after (right) subsequent steps of optimization and relaxation.

Goals

  • Comprehensive data visualization and precise definition of interactions
  • Fast computation methods for real time interactivity