Artifact (diagonal line/band)

[qrecursion]

OS: Linux - CentOS 8
Compiler: g++ (GCC) 8.5.0 20210514 (Red Hat 8.5.0-18)
STIR Version: v6.0.0 and v6.2.0
Data Format: Listmode-SAFIR convention
Lm to projdata Tool: lm_to_projdata
Data Simulation Tool: Geant4_v10.7

Scanner Geometry:
The scanner has 8 detector blocks transaxially and 2 detector blocks axially. There is only one layer. Each block detector consists of 35 crystals in each direction and so each detector block is an array of 35x35 crystals. Between each crystal is a gap of 0.1 mm and each crystal dimension is 1.2x1.2x13 mm^3. The scanner inner radius is 56 mm (please see the screenshots). We do not have any DOI.
Crystal material: LYSO
detector block material: Teflon
Isotope: F18

Artifact Description:
As you can see in the attached screenshots, the artifact is a diagonal line that appears at slices 68, 69, 70, 71, 72 out of 139 slices. The artifact is not present in other slices like 45 or 89.

Additional Information:

1. There is a relation between the radius of the scanner/source the artifact appearance. For example, if the source radius is 12mm then the artifact does not appear. Or, if the radius of the scanner is more than 59 (the gap), then even if the source radius is 15mm, the artifact does not appear.
2. There is no symmetry in the artefact, it appears diagonally in one direction and with a specific angle.
3. Setting "use the actual detector size" to 1, or "map model" to multiplicative, make the artifact looks faint. however, it still remains there.
4. The artifact is still present when parallelproj projector is used.
5. I increased the size of the crystals to 2.1x2.1x13 (and consequently, the size of detector blocks) and left everything else unchanged. But the artifact was present.
6. I used 12 detector blocks transaxially with each detector block contains 18 crystals (1.2x1.2x13) in each direction and the scanner radius was around 45, however, the artifact was present.
7. I changed almost all the parameters in OSMAPOSL parameter file, but it did not changed any thing regarding this artifact.

A change that removes the artifact:
When I changed my simulation setup in Geant4 so that instead of two blocks, the scanner is consist of one block axially but the detector block is now an array of 35x70. Leaving everything else the same, there is no artifact in the image anymore.

Please let me know if you need more information from me.

All the best,
Seyyed

Attachment:

1. Artifact (Screenshots are from two different data but the scanner is the same.)
   
   

2. Scanner
   
   

3. Parameter files
   Crystal_map.txt
   lm_to_projdata.txt
   OSMAPOSL.txt