Multi-b-value diffusion magnetic resonance imaging optimization method based on signal-to-noise ratio weighting

The invention discloses a multi-b-value diffusion magnetic resonance imaging optimization method based on signal-to-noise ratio weighting. The method comprises the following steps of through multi-b-value weight data, establishing a fitting model; through the multi-b-value weight data, estimating a signal to noise ratio; according to the signal to noise ratio, calculating a residual error correction weight alpha i and establishing a fitting model initial value; according to the residual error correction weight alpha i, calculating a fitting model residual error correction value; if the fitting model residual error correction value accords with a convergence condition, through an optimal solution of the fitting model, calculating and increasing stability; otherwise, returning to a step4 and continuously seeking the optimal solution of the fitting model. In the method, aiming at multi-b-value diffusion weighting, the fitting model is constructed; through optimization of a least square algorithm, stability of model optimization searching solution is increased so that estimation of a parameter image in MRI imaging is stable; and generation of a singular value or a meaningless value is reduced so as to further promote development and application of a MRI imaging technology.