H. Elser et al., SEMIQUANTITATIVE TC-99M-HMPAO-SPECT AND D EMENTIA OF THE ALZHEIMER-TYPE - INFLUENCE OF THE SELECTION OF RECONSTRUCTION-FILTER AND REFERENCE-REGION, Nuklearmedizin, 35(6), 1996, pp. 243-250
Aim: The aim of the study was to clarify, whether the selection of the
reference-region and/or the reconstruction-filter influences the resu
lt of the semiquantitative analysis of a Tc-99m-HMPAO-SPECT that was c
onducted to diagnose the dementia of the Alzheimer type (DAT). Methods
: A group of 19 DAT-patients according to the criteria of NINCDS-ADRDA
and DSM-III-R was examined together with a comparison group (n = 14)
with normal cerebral perfusion. Three reference-regions (cerebellum, w
hole slice, occipital cortex) and four reconstruction-filters (Hanning
f(c) = 0.7 and 1.0 Nyquist; Butterworth (n = 8); f(c) = 0.5 and 0.9 N
yquist) were applied to twelve standardized regions of interest (per p
atient) respectively. The data was evaluated through a ROC-analysis. R
esults: II has been showed, that the bilateral parieto-temporal perfus
ion reduction as a characteristic of DAT depends on the filters and re
ference-regions used. The most secure separation of both groups of pat
ients was obtained through a Butterworth-filter (n = 8; f(c) = 0.5) in
combination with the cerebellum as reference-region. Conclusion: The
selection of the reference-region and the reconstruction-filter has an
important influence on the results of a semiquantitative analysis. Th
erefore standardisation in dependency on the actual questioning is nec
essary.