Optimizing clinical DaTscan imaging

Objectives DaTscan (Ioflupane I-123) is a radiopharmaceutical used to differentiate Parkinson’s disease from disorders presenting similar symptoms. Current clinical imaging practices are largely based on research protocols and studies conducted prior to DaTscan’s clinical availability, including the use of low-energy, high-resolution collimators (LEHR) and detector rotations of 360°. The use of medium energy collimators (ME) has been reported to improve image quality in other imaging studies utilizing I-123 (i.e. MIBG) by decreasing septal penetration and down-scatter from I-123’s high-energy gamma emissions. The objective of this study is to examine the optimization of current clinical DaTscan imaging parameters.

Methods A phantom was constructed using a Hoffman brain phantom tank and fillable polypropylene spheres (PPS) grouped to simulate the uptake of DaTscan in the striatum. A known activity standard was imaged and compared with clinical DaTscan images to estimate a comparable PPS activity of 4.75 µCi/ml of I-123. Approximately 0.04 uCi/ml was loaded into the surrounding tank to simulate background. The phantom was imaged using the current clinical protocol; LEHR, detector rotation of 360° and 15sec/view. Subsequent scans were performed with the following: LEHR, 180°, 30sec/view; ME, 360°, 15sec/view; ME, 180°, 30sec/view. The current clinical reconstruction parameters (OSEM-3D; 8-iterations, 8-subsets, 6mm Gaussian post-filter and CT attenuation correction) were utilized for each acquisition.

Results Utilizing the current clinical reconstruction parameters, visual inspection determined ME imaging improved the contrast of the PPS to background, but with degraded resolution. Resolution also decreased slightly when considering a 360°, 15sec/view vs. 180°, 30sec/view for both collimators.

Conclusions Maintaining current clinical reconstruction parameters, introduction of ME alone did not improve overall image quality. However, further investigation is warranted as a limited evaluation of modified reconstruction parameters for the ME data sets indicated potential improvements in resolution, along with improved contrast, over LEHR.