[8] The convergence of the iterative migration/inversion

Again, the iterative formula for migration/inversion imaging is  

$$\delta \textbf{m}^{(k+1)} = \delta \textbf{m}^{(k)} - H^{k+1} \nabla f\left(\delta \textbf{m}^{(k)} \right) .$$ (93)

The choice of background parameters and the method for linearizing the propagators determine whether the cost function is (or approaches) a quadratic function. Therefore, the Born approximation should be replaced by De Wolf approximation or other more accurate approximations. On the other hand, to produce the image at the first iterative step with the so-called true-amplitude imaging approaches will accelerate the convergence of the iterative migration/inversion.