Abstract
BACKGROUND: Partial volume effects in atrophied areas should be taken into account when interpreting brain perfusion single photon emission computed tomography (SPECT) images of neurodegenerative diseases. To evaluate both perfusion and atrophy using brain SPECT alone, we developed a new technique applying tensor-based morphometry (TBM) to SPECT. METHODS: After linear spatial normalization of brain perfusion SPECT using 99mTc-ethyl cysteinate dimer (99mTc-ECD) to a Talairach space, high-dimension-warping was done using an original 99mTc-ECD template. Contraction map images calculated from Jacobian determinants and spatially normalized SPECT images using this high-dimension-warping were compared using statistical parametric mapping (SPM2) between two groups of 16 multiple system atrophy of the cerebellar type (MSA-C) patients and 73 age-matched normal controls. This comparison was also performed in conventionally warped SPECT images. RESULTS: SPM2 demonstrated statistically significant contraction indicating local atrophy and decreased perfusion in the whole cerebellum and pons of MSA-C patients as compared to normal controls. Higher significance for decreased perfusion in these areas was obtained in high-dimension-warping than in conventional warping, possibly due to sufficient spatial normalization to a 99mTc-ECD template in high-dimensional warping of severely atrophied cerebellum and pons. In the present high-dimension-warping, modification of tracer activity remained within 3% of the original tracer distribution. CONCLUSIONS: The present new technique applying TBM to brain SPECT provides information on both perfusion and atrophy at the same time thereby enhancing the role of brain perfusion SPECT.