Comparison of 2-dimensional and 3-dimensional cardiac Rb-82 PET studies

Most new PET scanners have the capability to collect data in 3-dimensional (3D) (septa removed) mode. This allows many more detected events at the cos t of increased random events and scatter. In the case of Rb-82 imaging, the injected dose might have to be limited to avoid saturating the scanner. We present a comparison of 2-dimensional (2D) and 3D data collection for Rb-8 2 cardiac studies using the ECAT EXACT scanner. Methods: Resting Rb-82 card iac studies were collected in 2D and 3D modes for 33 consecutive patients. Four experienced physicians rated the images to determine if the different acquisition methods would lead to different patient care. A separate quanti tative analysis was performed on data from multiple scans of a thoracic pha ntom filled to simulate cardiac and background radioactivity corresponding to Rb-82 injections between 37 and 1740 MBq. Results: The 2D and 3D studies were significantly different, with the image quality being poorer in the 3 D studies. The scanner collected data at near its maximal counting rate for either 1480-MBq 2D or 37-MBq 3D acquisitions. Because the data collection was counting rate limited in either mode, and there are more random and sca tter events in 3D mode, the 2D acquisitions resulted in more detected true events and a better signal-to-noise ratio. Conclusion: Cardiac Rb-82 studie s should be performed in 2D mode when using the ECAT EXACT scanner.