Impact of Acquisition Geometry, Image Processing, and Patient Size on Lesion Detection in Whole-Body 18F-FDG PET
03 medical and health sciences
Imaging, Three-Dimensional
0302 clinical medicine
Fluorodeoxyglucose F18
Neoplasms
Positron-Emission Tomography
Image Processing, Computer-Assisted
Humans
Whole Body Imaging
Radiopharmaceuticals
Body Mass Index
DOI:
10.2967/jnumed.108.007369
Publication Date:
2007-11-16T01:47:16Z
AUTHORS (6)
ABSTRACT
The aim of this work was to develop a rigorous evaluation methodology assess performance different acquisition and processing methods for variable patient sizes in the context lesion detection whole-body (18)F-FDG PET.Fifty-nine bed positions were acquired 32 patients 2-dimensional (2D) 3-dimensional (3D) modes 1-4 h after injection (740 MBq) using BGO PET scanner. Three spheres (1.0-, 1.3-, 1.6-cm diameter) containing (68)Ge also imaged separately air, at locations corresponding possible sites 2D 3D (590 targets per condition). Each position 7 min 6 corrected randoms delayed window subtraction (DWS) or variance reduction (RVR). Sphere sinograms attenuated attenuation map derived from transmission scan patient, scaling with identical factors ensure marginal detectability. Resulting reconstructed filtered backprojection (FBP) ordered-subsets expectation maximization (OSEM) without any scatter correction (FBP-NATS OSEM-NATS) FBP (FBP-ATT) attenuation-weighted OSEM (AWOSEM). processed identically Fourier rebinning. Next, volumes compared on basis 3-channel Hotelling observer (CHO-SNR [SNR is signal-to-noise ratio]) detecting presence sphere unknown size an anatomic background while modeling noise. noise equivalent count (NEC) rate computed 3 phantoms (40, 60, 95 kg) SNR.3D imaging yielded better detectability than (P < 0.025, 2-tailed paired t test) normal (body mass index [BMI] = 31). However, large (BMI > 31), as deteriorated 0.05). similar results sizes. CHO-SNR 40% greater AWOSEM, FBP-ATT, FBPNAT 0.05), AWOSEM significantly did FBP. In all patients, RVR systematic improvement over DWS both 3D. radicalNEC characterized by behavior that SNR(CHO) phantom considered study.
SUPPLEMENTAL MATERIAL
Coming soon ....
REFERENCES (0)
CITATIONS (33)
EXTERNAL LINKS
PlumX Metrics
RECOMMENDATIONS
FAIR ASSESSMENT
Coming soon ....
JUPYTER LAB
Coming soon ....