A method for magnetic resonance imaging (MRI) comprises applying a consecutive series of MRI sequences to a target volume (V) according to experimental settings (TR, α, β). A discrete sequence of transient response signals (Sn, Sn+1, Sn+2) is measured and fitted to a fit function (F) that is continuously dependent on a sequence number (n) of the respective MRI sequence (Pn) and corresponding response signal (Sn). A shape of the fit function is determined according to an analytically modelled evolution by the experimental parameters (TR, α, β) as well as variable intrinsic parameters (r, λ3, φ, δ) to be fitted. For example, the model is based on an equivalent harmonic oscillator. The intrinsic parameters of the fit function can be related to the intrinsic properties (PD, T1, T2) of the spin systems and used for imaging the target volume (V). Various optimizations of contrast can be achieved by tuning the experimental settings according to the model.