Improving the reliability of T2 measurement in magnetic resonance imaging

Abstract

Magnetic resonance imaging (MRI) is a versatile technique broadly utilized in research and clinic. Among the information available from MRI measurements, transversal relaxation time (T2) is a basic but critical one to reflect the microstructure and microenvironment at the molecular level. A popular method to estimate T2 is employing a series of evolution time (TE) values and, thereafter, fitting with the exponential model (termed as T2 fitting). Despite of the popularity in using T2 fitting, there is a lack of knowledge how related experimental parameters, such as signal-to-noise (SNR), number of TE utilized, dynamic range of TE, and repetition number of each TE, affect the measurement reproducibility. In this study, we performed Monte Carlo simulations to investigate how SNR, TE number, TE range, and repetition number influence the reliability of T2 measurement, which was indexed by coefficient of variations. Good reliability with [≤]5.0% variation can be achieved when the normalized noise level is below 2.2%. Optimal TE range for measuring T2 is related to the T2 under evaluation. TE number and repetition number can be increased to reduce measurement variations.