A Poisson resampling method for simulating reduced counts in nuclear medicine images.

Area of Science:

• Medical Imaging
• Nuclear Medicine Technology
• Image Processing

Background:

• Modern nuclear medicine computers utilize software for resolution recovery to enhance low-count images.
• These advanced techniques aim to provide equivalent clinical information compared to full-count images.
• Reducing counts in nuclear medicine imaging allows for lower administered activity or shorter scan times.

Purpose of the Study:

• To introduce a method for accurately simulating reduced-count nuclear medicine images.
• To address the inadequacy of simple division/rounding methods for maintaining correct counting statistics.
• To present manufacturer-independent software for retrospective simulation of reduced-count nuclear medicine data.

Main Methods:

• Applying Poisson resampling to original raw nuclear medicine data.
• Developing software capable of generating simulated reduced-count images.
• Validating the maintenance of appropriate counting statistics in simulated images.

Main Results:

• Poisson resampling accurately simulates reduced-count nuclear medicine images.
• The developed software provides manufacturer-independent retrospective simulation.
• Simulated images maintain appropriate counting statistics, unlike simpler methods.

Conclusions:

• Poisson resampling is a valid technique for simulating reduced-count nuclear medicine images.
• The developed software enables departments to evaluate reduced-count imaging protocols accurately.
• Accurate simulation supports optimized nuclear medicine workflows with reduced radiation exposure or acquisition time.