Resolution-recovery-embedded image reconstruction for a high-resolution animal SPECT system

Area of Science:

• Medical Imaging
• Nuclear Medicine
• Preclinical Imaging

Background:

• Dedicated dual-head gamma cameras are crucial for small-animal imaging in research.
• Existing systems often face limitations in achieving high resolution for detailed murine model visualization.
• Image reconstruction algorithms play a vital role in optimizing the performance of SPECT systems.

Purpose of the Study:

• To develop and evaluate a resolution-recovery-embedded image reconstruction code for a novel small-animal High-Resolution SPECT (HiReSPECT) system.
• To assess the impact of the developed algorithm on image resolution and noise performance.
• To validate the system and reconstruction method using phantoms and actual mouse scans.

Main Methods:

• Measured collimator-detector response functions (CDRFs) using capillary sources and fitted them with Gaussian functions.
• Created a continuous map of CDRFs by applying linear interpolation to Gaussian standard deviation values.
• Employed a rotation-based maximum-likelihood expectation maximization (MLEM) algorithm with a fast rotation method for image reconstruction.

Main Results:

• The resolution-recovery algorithm significantly improved image resolution across the field of view (FOV).
• Full-width at half-maximum (FWHM) resolution improved from over 2 mm (up to 2.5 mm) to below 1.8 mm everywhere within the FOV.
• Noise performance remained acceptable, with standard deviations of 6.6% and 8.3% without and with resolution recovery, respectively.

Conclusions:

• The developed resolution-recovery-embedded reconstruction code effectively enhances the spatial resolution of the HiReSPECT system.
• This advancement provides superior imaging capabilities for small animal studies, particularly in murine models.
• The improved resolution and acceptable noise levels support the utility of this system for preclinical research applications.