Generation of MR-Based Attenuation Correction Map of PET Images in the Brain Employing Joint Segmentation of Skull and Soft-Tissue from Single Short-TE MR Imaging Modality

Abstract

Recently introduced PET/MRI scanners present significant advantages in comparison with PET/CT, including better soft-tissue contrast, lower radiation dose, and truly simultaneous imaging capabilities. However, the lack of an accurate method for generation of MR-based attenuation map (μμ-map) at 511 keV is hampering further development and wider acceptance of this technology. Here, we present a new method for the MR-based attenuation correction map (μμ-map), employing a proposed short echo-time (STE) MR imaging technique along with the nearly automatic segmentation. This method repeatedly applies active contours inhomogeneity correction, multi-class spatial fuzzy clustering (SFCM), followed by shape analysis, to classify the images into cortical bone, air, and soft tissue classes. The proposed segmentation method returned sensitivity of 81 % for cortical bone and above 90 % for soft tissue and air. These results suggest that this technique is accurate, efficient, and robust for discriminating bony structures from the neighboring air and soft tissue in STE-MR images, which is suitable for generating MR-based μμ-maps.