Ray Tracing Algorithm

[saxu3829]

Hi,

I looked through the code and referenced papers on the ray tracing algorithm (Siggel et al. (2011), Siddon et al. (1984)). If I understood correctly, the algorithm only uses the density at the target point of the ray at each bixel. There is no calculation of the mean density along the path (between two bixels in the same ray), right?

Thank you in advance!
Jessica

[wahln]

I don't exactly understand what you mean.

The ray tracing algorithm itself obtains, for each voxel a ray traverses, the corresponding index, intersection length, density (or any cube value you want like rSP or rED).

We then use the ray-tracing algorithm in different ways:

• For spot positioning, we find the WEPL as cumulative sum over all rSP values at the target exit points by tracing on the central ray (sourcePoint to targetPoint of each ray). Then, up to the WEPL at the target entry point, we distribute spots according to their peakPos stored in the machine data.
• For dose calculation, we trace a large number of rays through the rSP cube to give every voxel a WEPL relative to the source. This value is then used during pencil-beam dose calculation as current depth when interpolating the pencil-beam kernels.

The ray.target_point is, by the way, only a helper entry placing the "end" of a ray opposite of the source.