Ziel/Aim Visual interpretation of amyloid-ß (Aß) PET is challenging in 10-20Ĩ % of the cases,
often due to brain atrophy. Co-registration with individual T1 w-MRI is useful in
these cases, but MRI is not always available. This study tested co-registration of
the Aß PET image with an early uptake image to support interpretation of Aß PET.
Methodik/Methods 31 patients were included retrospectively: 6 patients with florbetaben (FBB) PET
including an early uptake image from clinical routine and 25 ADNI MCI patients with
FBB PET and FDG PET (as replacement of early FBB uptake image). Both PET images were
co-registered to the individual high-resolution MRI. Cortical tracer uptake in Aß
PET was interpreted (Aß-negative or Aß-positive) in 4 ROIs: temporal, frontal, parietal
cortex, posterior cingulate cortex/precuneus (PC2). Visual interpretation was performed
by a nuclear medicine resident in the following settings (randomized order): (A) late
Aß image alone, (B) late and early Aß image combined, (C) late Aß image and MRI combined.
Certainty about the interpretation was scored from 0 (very uncertain) to 5 (very certain).
Ergebnisse/Results 52 % of the patients were interpreted Aß-positive (setting A). The rate of discrepancy
between setting B and setting A was 13 %, both on patient and ROI base. The rate
of discrepancy was smallest in PC2 (6 %) and largest in the parietal cortex (23 %).
The certainty of the binary ROI Aß interpretation was 3.74 ± 0.95, 4.59 ± 0.60, and
4.26 ± 0.82 in setting A, B, and C, respectively (all paired t-test p < 0.0005). The
increase in certainty in setting B relative to setting A was largest in temporal cortex
and PC2.
Schlussfolgerungen/Conclusions These findings suggest that co-registration with an early uptake image supports visual
interpretation of Aß PET. The small improvement of certainty by co-registration with
the early PET image compared to individual MRI is due to better matching of spatial
resolution between the two PET images.
