Technical note: Partitioning of gated single photon emission computed tomography raw data for protocols optimization

Area of Science:

• Nuclear Medicine
• Medical Imaging

Background:

• Optimizing Single-Photon Emission Computed Tomography (SPECT) imaging is crucial but faces challenges like throughput, physiological bias, and patient comfort.
• Existing methods for simulating reduced SPECT data are often vendor-specific, limiting broader research applications.

Purpose of the Study:

• To evaluate a vendor-independent algorithm for simulating SPECT image acquisitions with lower photon counts.
• To assess the feasibility of generating reduced count datasets from existing SPECT raw data.

Main Methods:

• Developed a novel algorithm to recombine Electrocardiogram (ECG)-gated raw SPECT data into acquisitions with reduced counts.
• Tested the algorithm using phantom SPECT acquisitions synchronized with an ECG simulator.
• Reconstructed simulated datasets using a resolution recovery algorithm and assessed the Summed Stress Score (SSS) by human and automated readers.

Main Results:

• The algorithm successfully generated datasets simulating various lower counting statistics, effectively mimicking multiple examinations.
• Achieved low error rates: 5%-10% for ungated simulations and 0% for gated simulations between expected and simulated counts.
• Demonstrated the algorithm's capability to create reduced count SPECT data from single-gated raw data.

Conclusions:

• The developed vendor-independent algorithm effectively simulates lower counting statistics in SPECT imaging.
• This methodology offers a valuable tool for SPECT research, enabling studies that aim to reduce radiotracer doses or shorten scan durations.
• Facilitates optimization of SPECT acquisition protocols without requiring specialized vendor software.