The dij struct
The dij struct holds pre calculated dose influence data for inverse planning. The individual fields contain the following information
dij.numOfBeams
Specifies the number of beams used for dose calculation and consequently inverse planning.
dij.numOfVoxels
The total number of voxels in the entire ct cube.
dij.resolution
The resolution of an individual voxel in the ct cube in [mm] in x-y, y-, and z-direction.
dij.numOfRaysPerBeam
Specifies the number of rays per beam. For photons, this number also correponds to the number of bixels per beam. For particles however, it is also possible to have multiple spot positions with different energies at the same lateral spot position.
dij.totalNumOfRays
Specifies the total number of all rays, i.e. dij.totalNumOfRays = sum(dij.numOfRaysPerBeam).
dij.totalNumOfBixels
Specifies the total number of bixels over all beams.
dij.bixelNum
Lists the bixel number in an individual beam for all columns in the precomputed influence data. Together with dij.rayNum and dij.beamNum this information facilitates an easy assignment of columns of the influence data to the stf struct.
dij.rayNum
Lists the ray number in an individual beam for all columns in the precomputed influence data. Together with dij.bixelNum and dij.beamNum this information facilitates an easy assignment of columns of the influence data to the stf struct
dij.beamNum
Lists the beam number for all columns in the precomputed influence data. Together with dij.bixelNum and dij.rayNum this information facilitates an easy assignment of columns of the influence data to the stf struct
dij.physicalDose
Pre-computed dose influence matrix with dij.numOfVoxels rows and dij.totalNumOfBixels columns. This matrix specifies the dose contribution from every bixel to every voxel, stored with matlab's built-in double preciion sparse matrix format.
dij.mAlphaDose
Pre-computed alpha*dose matrix with dij.numOfVoxels rows and dij.totalNumOfBixels columns, stored with matlab's built-in double preciion sparse matrix format. This matrix is only computed for biological optimization, where this information is required to compute dose-averaged alpha cubes which are in turn required for three-dimensional RBE modeling.
dij.mSqrtBetaDose
Pre-computed ssqrt(beta)*dose matrix with dij.numOfVoxels rows and dij.totalNumOfBixels columns, stored with matlab's built-in double preciion sparse matrix format. This matrix is only computed for biological optimization, where this information is required to compute dose-averaged alpha cubes which are in turn required for three-dimensional RBE modeling.
Screenshot of the dij-struct:
