Converting list mode of events to STIR (interfile) sinograms

This implementation concerns data, simulated in GATE for the cylindrical scanner in the framework of Jagiellonian Positron Emission Tomography (J-PET) project.

Prerequisites

• STIR 3.0 (for registered users) or developers’ version
• SAFIR, build with STIR 3.0 (but already integrated with developers’ version, so no need to install).

Easy installation:

$ source install.sh

To load necessary paths to variables, do before every use:

$ source setup.sh

The file setup.sh will be created after running easy installation.

Easy use:

Copy all data files to working_dir/INPUT_DATA and

$ cd working_dir
$ ./one_click_reconstruction.sh

Step-by-step build (use if easy install fails)::

$ cd /directory-with-repository-files/src/
$ mkdir -p build
$ cd build
$ cmake ../
$ make

run this outside this shell (inside build directory)

$ export PATH=$PWD:$PATH

Usage examples:

The executables are (partially tested):

• GenerateCrystalMapJPET
• ConvertToMUPET
• GenerateBlurredMUPET

$ GenerateCrystalMapJPET -p parameters_file.par -o crystal_map_file.txt
$ ConvertToMUPET -i input_file -p parameters_file.par -o output_file.clm.safir
$ GenerateBlurredMUPET -i input_file [-p parameters_file.par]

Please, refer to SAFIR example for further instructions.

For the latter executable GenerateBlurredMUPET, it is optional to set geometry file with -p key. If not given, the script would try to generate .par file by parsing the very name of the input (for example, "D85_1lay_L020_4mm_10_00_07"). But even if the name could not be parsed, the program would search for the default parameters file "JPETGeometry.par" as the last possible option.

Consequently, GenerateBlurredMUPET would add new files, corresponding to three photomultipliers that reflect different smearing of the data, in both listmode and SAFIR formats.

Note! At the moment, the "smeared" listmode data comprise only pairs of Cartesian coordinates with times of hits, cutting out the rest (energy, type of coincidence etc).

Note! Smearing hit positions along axial (Z) direction would probably leave some events outside the scanner. By default, such events are thrown away, reducing their total number. Alternatively, hit positions could be coerced to the edges of the scanner (setting coerceEdges to true in JPETHit::calibrate() and JPETEvent::calibrate() functions).