Comparative assessment of statistical brain MR image segmentation algorithms and their impact on partial volume correction in PET

Novel Quantitative Techniques in Hybrid (PET-MR) Imaging of Brain Tumors

“…39 In brain imaging, MR imaging-guided PET image reconstruction was reported to outperform CTguided reconstruction owing to the high soft tissue contrast provided by MR and the accuracy obtained using sophisticated brain MR imaging segmentation procedures. 40 The quantitative accuracy of PET activity concentration estimates for sources having dimensions less than twice the system's spatial resolution is limited because the counts in smaller volumes are spread over a larger volume than the physical size of the object owing to the limited spatial resolution of the imaging system. This phenomenon is referred to as the partial volume effect (PVE) and can be corrected using one of the various strategies developed for this purpose.…”

“…The accuracy of MR imaging-guided PVE correction in PET largely depends on the accuracy achieved by the PET-MR imaging coregistration procedure, which is improved by using simultaneous hybrid PET-MR imaging systems. Zaidi and colleagues 40 evaluated the impact of brain MR image segmentation methods on PET partial volume correction in 18 F-FDG and 18 F-L-dihydroxyphenylalanine ( 18 F-DOPA) brain PET imaging. The results indicated that a careful choice of the segmentation algorithm should be made while using geometric transfer matrix-based partial volume corrections in brain PET.…”

Partial Volume Correction Strategies in PET

“…Overall, the accuracy of the corrected estimates seems to be impacted more by the success of the segmentation of the structural information provided by MR imaging than by the influence of image co-registration. Zaidi and colleagues [86] investigated the impact of brain MR image segmentation on PVC. Three algorithms were compared: the first, bundled in the Statistical Parametric Mapping package [87], and the second, the Expectation Maximization Segmentation algorithm [88], incorporate a priori probability images derived from MR images of a large number of subjects.…”

Deconvolution-based partial volume correction in Raclopride-PET and Monte Carlo comparison to MR-based method

“…The effects of mis-registration and missegmentation in MR-based PV correction have been previously studied by Baete et al (2004); Frouin et al (2002); Meltzer et al (1999); Quarantelli et al (2004) and Zaidi et al (2006). A majority of the above mentioned works targeted 18 F-FDG-PET imaging of cortical regions.…”