Optimal Pulse Sequences for Magnetic Resonance Imaging-Computing Accurate T1, T2, and Proton Density Images

TLDR

Proton MRI images depend on proton density, T1, T2, and scan parameters, and systematic measurement errors vary with pulse sequence, causing large deviations between objective images computed by different methods. The study aimed to improve resolution and reduce errors in computed T1, T2, and proton density images across representative human tissues. The authors tested multiple scan sequences and evaluated methods to achieve this improvement. The combination of IR3SE and SR4SE sequences produced more accurate computed images for a fixed scan time, and adopting it as a standard enables comparison of clinical results across researchers.

Abstract

Proton nuclear magnetic resonance images (proton MRI) are functions of not only proton density (ρ), spin lattice relaxation time (T1), and spin-spin relaxation time (T2) but also MRI scan parameters, so the images differ for different pulse sequences and scan parameters. Systematic measuring errors also vary depending on the type of pulse sequence and the scan parameters. "Pure," objective, T1, T2, and proton density images can be computed from several different Proton MRI, however, systematic measuring errors can cause large deviations between "pure" images obtained using different methods. We tested several scan sequences, and investigated methods to improve the resolution and reduce the errors over the range of T1, T2, and proton density values encountered in representative human tissues. We found that, for given scan time, the combination of inversion recovery 3 spin echo (IR3SE) and saturation recovery 4 spin echo (SR4SE) sequences gave more accurate computed images than other comparable methods tested. It follows, then, that adopting such a sequence as "standard" allows meaningful comparison of clinical results obtained by different researchers.