[WIP] Particle Reconstruction with Grid Cells

This example uses the rpt_cell_reconstruction_3d application to find particle positions of a radioactive particle using experimental counts data provided by 6 detectors. It uses grid cells for reconstruction, which means, it aims to find the cell which, particle exist in and assigns the center position of the "final best" cell to the particle position for each recorded photon count in the particle trajectory.

Parameter file

In the subsection “rpt parameters”, we define common parameters for the RPT simulation which are described in the RPT parameters subsection in the RPT Parameters documentation. The reactor height is an important parameters for this feature since all the dimension of the reactor are needed for the grid.

# --------------------------------------------------
# RPT Monte Carlo technique
#---------------------------------------------------
subsection rpt parameters
    set verbosity                        = verbose
    set monte carlo iteration            = 1000
    set random number seed               = 0
    set reactor height                   = 0.3
    set reactor radius       		 = 0.1
    set peak-to-total ratio  		 = 0.4
    set sampling time        		 = 1
    set gamma-rays emitted        	 = 2
    set attenuation coefficient detector = 21.477
    set dead time       		 = 1e-5
    set activity  			 = 2e6
    set attenuation coefficient reactor  = 10
end

In the subsection “reconstruction parameters”, we specify the coarse mesh level of the reactor grid and the refinement level, the reconstruction counts file provides the measured counts by all detectors. As in this example the measured counts and the real related positions are known, by activating the analysis option we are allowed to track the precision of the algorithm to find out if the final reconstructed cell is the same cell than the known position. Also, it assigns a status to the final cell giving information about the use of cost function or not as the incapacity to find a cell at refined level or not.

Flag of the best cells after analysis of reconstructed positions vs real positions (many status can be associated to a cell) are explained as follows :

• right_cell : The reconstructed position is in the same cell than the real position
• cost_function : There was many cell candidates at refined mesh and a cost function was needed to get the best cell
• parent_cell : The best cell is not at the highest refined level
• wrong_cell : The reconstructed position is not in the same cell than the real position

Depending of flags associated to the final best cell, it generates a status :

• Status 0 : The final cell for the reconstructed position contains the real position
• Status 1 : The final cell for the reconstructed position contains the real position but required the cost function use
• Status 2 : The final cell for the reconstructed position contains the real position but is not at the refined level
• Status 3 : The final cell for the reconstructed position contains the real position but required the cost function use and is not at the refined level
• Status 4 : The final cell for the reconstructed position does not contain the real position
• Status 5 : The final cell for the reconstructed position does not contain the real position and required the cost function use
• Status 6 : The final cell for the reconstructed position does not contain the real position and is not at the refined level
• Status 7 : The final cell for the reconstructed position does not contain the real position, required the cost function use and is not at the refined level

The reconstruction parameters are described in the Reconstruction section of the RPT Parameters documentation.

# --------------------------------------------------
# Particle reconstruction
#---------------------------------------------------
subsection reconstruction
    set coarse mesh level                       = 1
    set refinement                              = 6
    set reconstruction counts file              = counts.reconstruction
    set export reconstruction positions file    = reconstruction.csv
    set analyse positions                       = true
    set known positions file                    = positions.known
end

In the subsection “detector parameters”, we specify the file that contains the position of the 6 detectors face center and the position of a point inside a detector on its axis. The detector parameters are described in the Detector section of the RPT Parameters documentation.

#---------------------------------------------------
# Detector parameters
#---------------------------------------------------
    set detector positions file          = positions.detector
    set radius       			 = 0.0381 
    set length                           = 0.0762
end

Results

...