Accuracy assessment of fast dynamic volumetric measurements with a three-headed SPECT system

Abstract

185

Objectives: The clinical impact of SPECT measurements of dopamine receptor density is hampered by the low capability of SPECT to recover activity concentration ratios (ACRs) [1]. In [2] we proposed a method for acquiring and processing data when three-headed systems are used in DSPECT. We showed that when few regions of interest (ROI) are concerned, three views data sets may be sufficient to reconstruct ROI contents directly from projections. This phantom study evaluates the accuracy of our DSPECT method to recover ACRs in selected brain regions.

Methods: An Alderson striatal phantom (www.rsdphantoms.com) was filled with striatal/background ratios of 6:1, 9:1, 28:1 (this is the range of the expected ACRs with dopamine system tracers). The temporal sampling of the DSPECT was settled as to comply with the count statistics of clinical SPECT studies with dopaminergic tracers. A total of 120 time points were acquired. The gamma-camera used for this study was an Irix (Philips). Briefly, the striatal and background ROIs were drawn on the basis of an X-ray CT of the phantom. The coregistration matrix between X-ray-CT and SPECT was obtained with Neurostat (http://128.95.159.66/~Download/) with the use of the images of a conventional SPECT acquisition on a 360 degree rotation. The ACRs of these ROIs were derived both from the DSPECT acquisition and from the conventional SPECT EM reconstruction. The influence of a segmentation error up to 20% of the striatal volume was also evaluated.

Results: The ACRs obtained with EM reconstruction were 3.2, 4.9 and 15.9 respectively at the points of 6:1, 9:1 and 28:1. At the same points with DSPECT the ACRs were 6.2+/-0.6, 9.0+/-0.9 and 27.2+/-2.6. The segmentation error gave greater changes in DSPECT ACRs (up to 15% for a 20% error) than in EM ACR values (up to 5%).

Conclusions: Fast DSPECT is feasible and accurate in the clinical examination of the dopaminergic system. This holds on a wide range of ACR values. This approach can be fruitful in disease progression monitoring, in developing new radiopharmaceuticals for SPECT imaging of the dopaminergic system and in clinical trials of dopaminergic drugs. [1] M. Soret, P. M. Koulibaly et al.: J. Nucl. Med., 44(7): 1184-1193, 2003. [2] E. Vanzi, A. R. Formiconi et al.: IEEE Trans. Med. Imag., 23(3): 363-373, 2004.