Tamino Huxohl
mu-map

Repository



Mu Map
Repository for the code to replicate the paper of Shi et al. with our SPECT/CT Scanner.

Differences


Authors
Ours


SPECT/CT Scanner
GE NM/CT 850 SPECT/CT
Siemens Symbia Intevo 16 SPECT/CT


Scan Arc
180°
208°


Angles
60
34


Energy Window Lower
114-126 keV
110-131 keV


Energy Window Upper
126-155 keV
131-152 keV


Dose
15 mCi
?


Reconstruction
OSEM (5 iterations, 4 subsets)
Siemens Conjugate Gradient


CT Energy
120 kVp
?


Attenuation Map Shape
(25-35)x64x64
?


Reconstruction Shape
64x64x64
51x128x128


Voxel Size
6.8x6.8x6.8 mm³
4.8x4.8x4.8 mm³


Open Questions

References


Shi et al. "Deep learning-based attenuation map generation for myocardial perfusion SPECT". 2020 In: European Journal of Nuclear Medicine and Molecular Imaging 47.10

DOI: 10.1007/s00259-020-04746-6


Install
Install libraries for tomographic reconstruction:

NiftyRec

mkdir libs
cd libs
git clone https://github.com/TomographyLab/NiftyRec
cd NiftyRec
git checkout 5329496

cp README.md README.txt - this file is needed by cmake
mkdir build
cd build

ccmake .. - turn of the volume renderer
make -j 8
make package
sudo dpkg -i NiftyRec-3.1.0-Linux-x86_64.deb


TomoLab
Make sure your python virtual environment is active for this.

mkdir libs
cd libs
git clone https://github.com/TomographyLab/TomoLab
git checkout 86b9a58