Fully 4D list-mode reconstruction applied to respiratory-gated PET scans

Area of Science:

• Nuclear Medicine
• Medical Imaging
• Image Reconstruction

Background:

• (18)F-fluoro-deoxy-glucose ((18)F-FDG) positron emission tomography (PET) is crucial for diagnosing non-small cell lung cancer.
• Respiratory motion during PET scans causes image blurring, degrading diagnostic quality.
• Current respiratory gating techniques increase image noise or scan duration.

Purpose of the Study:

• To evaluate a novel four-dimensional (4D) image reconstruction algorithm for PET.
• To assess its ability to combine gated data while preserving motion deblurring.
• To compare its performance against standard reconstruction methods.

Main Methods:

• Comparison of a 4D reconstruction algorithm with classic ordered subset expectation maximization (OSEM) for gated and non-gated images.
• Evaluation using a moving NEMA IEC/2001 body phantom and Monte-Carlo simulations of the NCAT breathing phantom.
• Assessment of signal-to-noise ratio (SNR) and motion deblurring capabilities.

Main Results:

• The 4D reconstruction algorithm achieved performance comparable to non-gated PET in terms of SNR.
• It effectively preserved motion deblurring, unlike standard gated OSEM reconstruction.
• The 4D method demonstrated superior SNR compared to gated OSEM and temporal filtering techniques.

Conclusions:

• 4D image reconstruction is a promising technique for PET imaging.
• It offers improved SNR and reduced motion blur without extending acquisition times.
• This method enhances diagnostic accuracy for lung cancer detection using PET scans.