Fusion Algorithm in Magnetic Resonance/Positron Emission Tomography using Attenuation Correction based on Ultrashort Echo-Time Pulse Sequences

Abstract

Positron emission tomography (PET) images in magnetic resonance (MR)/PET fusion imaging systems are corrected for parts attenuated by gamma-rays by various MR pulse sequences. We aimed to study the quality of MR/PET images using an attenuation correction method based on ultrashort echo-time (UTE) pulse sequences. The proposed image-quality improvement algorithm was modeled as a convergence of the median-modified Wiener filter (MMWF) for noise reduction and Prewitt operator for emphasizing the edge area. By applying the proposed algorithm to MR/PET images, superior contrast to noise ratio and coefficient of variation values were obtained compared to those in the original image. The edge rise distance data of the proposed algorithm exhibited a very small difference from that of the original MR/PET image (difference: 2.21 %). In conclusion, we confirmed the applicability and usefulness of MMWF and Prewitt operator in UTE-based MR/PET imaging.