Attenuation Correction for Phantom Images Acquired in the Second Generation PETcoil Insert for PET/MR Brain Imaging

This paper proposes a geometry-based attenuation correction method for phantom PET scans using emission data alone. The method reconstructs the phantom image without attenuation and scatter correction, registers the image with the known phantom, and calculates the intersection of every line of response (LOR) with the registered phantom geometry. Knowing the attenuation per LOR allows us to manipulate the list-mode emission data accordingly and generate a list-mode data file with an enhanced number of coincidences per LOR according to the estimated attenuation.The approach yielded promising results when applied to a Hoffman brain phantom study using the second-generation PETcoil insert scanner. The intensity line profile generated by summing the signal intensity along the transverse and coronal axes looked more like the expected line profile without scatter correction, as expected. This approach can be easily generalized to any other phantoms and scanners.