Local validation of the use of Evolution for Bone for bone SPECT imaging

Area of Science:

• Nuclear Medicine
• Medical Imaging Analysis
• Radiology

Background:

• Single-photon emission computed tomography (SPECT) is a crucial imaging modality in nuclear medicine.
• Ordered subsets expectation maximization (OSEM) is a standard reconstruction algorithm for SPECT imaging.
• Image quality in SPECT can be affected by noise and resolution, particularly at lower photon counts.

Purpose of the Study:

• To validate the efficacy of the Evolution for Bone software in improving bone SPECT image quality.
• To compare images reconstructed with Evolution for Bone against standard OSEM reconstruction.
• To assess the impact of reduced photon counts on image quality using Evolution for Bone.

Main Methods:

• Retrospective review of 30 bone SPECT datasets (spine and pelvis).
• Simulated reduced photon counts (50%, 75%, 100%) using Poisson resampling.
• Applied Evolution for Bone with resolution recovery and compared to OSEM reconstructed images.
• Qualitative blinded assessment of noise, contrast, and resolution by two independent radiologists.

Main Results:

• Both observers noted improvements in resolution and noise with Evolution for Bone across all datasets (P=0.776 for resolution, P=0.007 for noise).
• Overall image quality was perceived as improved with Evolution for Bone compared to standard OSEM.
• Disagreement was noted in contrast assessment, though results were deemed understandable.

Conclusions:

• Evolution for Bone demonstrates potential for enhancing bone SPECT image quality, even with reduced photon counts.
• The software offers significant advantages over traditional OSEM reconstruction in terms of resolution and noise reduction.
• Implementation of Evolution for Bone could lead to reduced patient radiation doses and shorter waiting lists in clinical practice.