Partial volume effect correction in SPECT for striatal uptake measurements in patients with neurodegenerative diseases: impact upon patient classification.

PURPOSE: In single-photon emission computed tomography (SPECT) of the dopaminergic system, measurements of striatal uptake are useful for diagnosis and patient follow-up but are strongly biased by the partial volume effect (PVE). We studied whether PVE correction might improve patient classification based on binding potential (BP) measurements.
METHODS: Patients with a probable diagnosis of dementia with Lewy bodies (DLB, 10 patients) or Alzheimer's disease (AD, 13 patients) were studied by( 123)I-FP-CIT SPECT. SPECT images were reconstructed with and without PVE correction. Each patient SPECT scan was also simulated to obtain SPECT data whose characteristics were fully known. In addition, 17 SPECT scans were simulated with striatal uptake values mimicking pre-symptomatic cases of DLB.
RESULTS: Without PVE correction, mean putamen BP values were 2.9+/-0.4 and 0.9+/-0.2 for AD and DLB patients respectively, while with PVE correction, they were 8.6+/-1.5 and 1.9+/-0.5 respectively. All patients were properly identified as having AD or DLB when considering mean putamen BP measured on their real or simulated SPECT scan, with and without PVE correction. All 30 simulations mimicking pre-symptomatic DLB and AD patients were accurately classified with PVE correction, but without PVE correction 15 mean putamen BP values were in a range where AD and DLB could not be distinguished.
CONCLUSION: We conclude that putamen BP values measured without PVE correction can be used to differentiate probable DLB and AD due to the already severe reduction in dopamine transporter levels. PVE correction appeared useful for accurate differential diagnosis between AD and pre-symptomatic DLB.