Influence and compensation of truncation artifacts in MR-based attenuation correction in PET/MR

Abstract

Insights

Truncation artifacts in PET/MR imaging cause significant SUV underestimations in arms, but a new method (TC2) effectively minimizes these errors. Reliable truncation compensation is crucial for accurate PET imaging.

Area of Science:

  • Medical Imaging
  • Radiological Physics
  • Positron Emission Tomography

Background:

  • Magnetic resonance (MR)-based attenuation maps (MRMap) are used in positron emission tomography (PET)/MR imaging.
  • Truncation artifacts in MRMap can affect the accuracy of reconstructed PET image volumes.

Purpose of the Study:

  • To quantify the influence of MRMap truncation artifacts on PET image volumes.
  • To propose and evaluate a new method for minimizing these truncation artifacts.

Main Methods:

  • Compared PET/MR reconstructions with and without two truncation compensation (TC) methods (TC1, TC2).
  • Assessed MRMap quality and truncation extent visually and compared TC2 to transmission-based attenuation maps.
  • Evaluated the influence of truncation on standardized uptake values (SUV) in various regions.

Main Results:

  • TC2 showed fewer residual artifacts in arms compared to TC1.
  • TC2 slightly overestimated arm size but significantly reduced SUV underestimations.
  • SUV underestimations were <5.4% in the trunk but up to 46.5% in the arms without compensation.

Conclusions:

  • Truncation artifacts in MRMap lead to minor SUV underestimations in the trunk but severe ones in the arms.
  • The newly developed TC2 algorithm reliably compensates for truncation artifacts.
  • TC2 is straightforward to implement and adaptable to other PET/MR systems.