Task-based assessment of binned and list-mode SPECT systems

Area of Science:

• Nuclear Medicine
• Medical Imaging Physics
• Quantitative Analysis

Background:

• List-mode (LM) data in SPECT offers higher precision than binned data.
• Investigating the clinical utility of LM data's precision is crucial.
• The energy attribute of detected photons in LM data is a key area of interest.

Purpose of the Study:

• To quantitatively evaluate if processing SPECT data in LM format improves absolute quantification of region-of-interest (ROI) uptake.
• To compare LM data processing, including the energy attribute, against binned data processing.
• To assess performance in the context of DaTscan brain SPECT for Parkinson's disease imaging.

Main Methods:

• Utilized a synthetic phantom with a signal-known exactly/background-known-statistically (SKE/BKS) setup.
• Employed an ordered-subset expectation-maximization (OSEM) algorithm for image reconstruction from LM data.
• Included scatter-window data and the energy attribute of each LM event in reconstruction.
• Performed quantification tasks on reconstructed images using a realistic 2-D SPECT system simulation.

Main Results:

• LM data processing consistently demonstrated improved quantification performance compared to binned data across all evaluations.
• Inclusion of the energy attribute in LM data processing further enhanced performance.
• The study confirmed enhanced absolute quantification accuracy using LM SPECT data.

Conclusions:

• List-mode (LM) SPECT data, particularly when incorporating the energy attribute, offers superior performance for absolute quantification tasks.
• This improved precision in quantitative SPECT imaging has significant implications for clinical applications, such as Parkinson's disease diagnosis.
• LM data processing represents a valuable advancement for quantitative accuracy in nuclear medicine.