Improved quantitation and reproducibility in multi-PET/CT lung studies by combining CT information

Area of Science:

• Medical Imaging
• Nuclear Medicine
• Radiology

Background:

• Accurate attenuation maps are crucial for reliable PET parameter estimation.
• Respiratory motion and multi-scan protocols in PET/CT can introduce significant errors.
• Previous work highlighted errors from incorrect attenuation maps in dynamic PET lung imaging.

Purpose of the Study:

• To develop and evaluate a method for combining multi-CT data for attenuation correction in dynamic PET/CT studies.
• To assess the impact of this combined-CT method on the precision of kinetic parameter estimates in patients with idiopathic pulmonary fibrosis.

Main Methods:

• A novel method was developed to create a composite CT (combined-CT) by averaging information from multiple CT scans (cine-CT, snapshot CT).
• This combined-CT was applied for attenuation correction in dynamic 18F-FDG PET/CT scans of nine patients with idiopathic pulmonary fibrosis.
• XCAT simulations were used to initially assess the method's performance.

Main Results:

• XCAT simulations showed potential errors up to 60% in influx rate constants with conventional methods.
• Patient data revealed an average standard error of 53% in influx rate constant estimates without combined-CT.
• The use of combined-CT for attenuation correction reduced this standard error to within 5%.

Conclusions:

• Reconstructing multi-section PET/CT studies using combined-CTs significantly enhances the precision of parameter estimates compared to using individual CTs.
• This improved precision may lead to better differentiation between diseased and healthy lung tissue in PET/CT imaging.