3D spatial resolution map and sensitivity characterization of a dedicated cardiac CZT SPECT camera

Objectives A solid state CZT SPECT device (Discovery NM 530c, GE Healthcare) provides high resolution and sensitivity in a field-of-view (FOV) defined as a sphere 18 cm in diameter. Present study aims at determining the spatial resolution throughout this FOV and measuring the sensitivity and its linearity, as compared to conventional cardiac SPECT (CCS) (Ventri, GE Healthcare).

Methods A phantom comprised of a 2D array of 115, ~1.5 MBq ‘point’ sources of Tc^99m on a 2.5 cm × 2.5 cm grid aligned with the transverse plane was shifted throughout 25 locations in the FOV and data was acquired for 2 min. each. Standard manufacturer’s reconstruction algorithms were applied for both systems. Full-width half-maximum (FWHM) of each source was calculated along the 3 Cartesian axes and interpolated to yield 3D spatial resolution maps. Sensitivity measurements used 0.37, 0.74, and 1.48 MBq Tc^99m sources separated by 2 cm imaged at central locations for both devices. No attenuation correction was applied.

Results The average spatial resolution for the 2D array was 7.2±3.4 mm throughout the Discovery NM 530c FOV and 5.1±1.0 mm within 4 to 13 cm of the detector face, vs. 16.7±3.5 mm for CCS. The sensitivity ranged from 0.28 to 0.72 kcps/MBq farthest and closest to the detector respectively, compared to 0.14 kcps/MBq for CCS. The reconstructed activity was linear within experimental error (±8%).

Conclusions The best average spatial resolution, 5.1 mm FWHM, is found at 4-13 cm from the detector of the Discovery NM 530c. Throughout the FOV it is better than that of conventional cardiac SPECT systems by a factor of 2. The sensitivity is linear and a factor of 2-5 higher than CCS and depends on the location within the FOV.