Determination of optimal parameters for 3D single‐point macromolecular proton fraction mapping at 7T in healthy and demyelinated mouse brain

Purpose To determine optimal constrained tissue parameters and off‐resonance sequence parameters for single‐point macromolecular proton fraction (SP‐MPF) mapping based on a comprehensive quantitative magnetization transfer (qMT) protocol in healthy and demyelinated living mice at 7T. Methods Using 3D spoiled gradient echo‐based sequences, a comprehensive qMT protocol is performed by sampling the Z‐spectrum of mice brains, in vivo. Provided additional T 1 , and B 0 maps allow for the estimation of qMT tissue parameters, among which three will be constrained, namely the longitudinal and transverse relaxation characteristics of the free pool (R 1,f T 2,f ), the cross‐relaxation rate (R) and the bound pool transverse relaxation time (T 2,r ). Different sets of constrained parameters are investigated to reduce the bias between the SP‐MPF and its reference based on the comprehensive protocol. Results Based on a whole‐brain histogram analysis about the constrained parameters, the optimal experimental parameters that minimize the global bias between reference and SP‐MPF maps consist of a 600° and 6 kHz off‐resonance irradiation pulse. Following a Bland‐Altman analysis over regions of interest, optimal constrained parameters were R 1,f T 2,f = 0.0129, R = 26.5 s −1 , and T 2,r = 9.1 µs, yielding an overall MPF bias of 10 −4 (limits of agreement [−0.0068;0.0070]) and a relative variation of 0.64% ± 5.95% between the reference and the optimal single‐point method across all mice. Conclusion The necessity of estimating animal model‐ and field‐dependent constrained parameters was demonstrated. The single‐point MPF method can be reliably applied at 7T, as part of routine preclinical in vivo imaging protocol in mice.