Reducing respiratory motion artifacts in positron emission tomography through retrospective stacking

Area of Science:

• Medical Imaging
• Radiochemistry
• Image Processing

Background:

• Respiratory motion significantly degrades positron emission tomography (PET) image quality.
• Artifacts reduce lesion intensity, activity, and contrast-to-noise ratios (CNRs), impacting diagnostic accuracy.
• Current methods may require additional scan time or are less effective.

Purpose of the Study:

• To introduce and evaluate a novel algorithm, retrospective stacking (RS), for correcting respiratory motion artifacts in PET imaging.
• To assess the efficacy of RS in restoring lesion characteristics and improving image quality.
• To demonstrate the benefits of RS without extending patient scan duration.

Main Methods:

• Developed retrospective stacking (RS), an algorithm utilizing b-spline deformable image registration.
• Combined amplitude-binned PET data across the respiratory cycle into a single respiratory endpoint.
• Applied RS to a phantom model and clinical 18FDG-PET scans (pancreatic and liver patients).

Main Results:

• Retrospective stacking accurately restored lesion location and intensity profiles in phantom and patient data.
• RS achieved CNR improvements of up to threefold over gated images and fivefold over ungated data.
• The method effectively corrected for lesion motion and deformation, enhancing tumor visibility.

Conclusions:

• Retrospective stacking is an effective technique for mitigating respiratory motion artifacts in PET imaging.
• RS significantly improves tumor visibility and reduces background noise, leading to better diagnostic potential.
• This approach offers substantial image quality enhancement without the need for additional acquisition time.