In this article, a modified microwave imaging method for stroke diagnosis is proposed. This method is based on introducing a suitable initial guess for the distorted Born iterative method (DBIM), which leads to improving the quality of DBIM images. To this end, as a further novelty of the article, we investigate the dependence of the Born iterative method (BIM) and DBIM on the initial guess. We demonstrate that the final result of the BIM is independent of the initial guess; however, this is not true for DBIM. On this basis, our proposed method has three steps. First, the image of the brain is reconstructed by BIM with an arbitrary initial guess. Then, this image is compared with a set of MRI-based brain images using the structural similarity index measure (SSIM) criterion to choose a suitable initial guess for DBIM. Finally, this initial guess is applied to DBIM to achieve an improved brain image. The performance of the proposed method is compared to that of BIM and DBIM with different initial guesses. The comparison results show that the accuracy and quality of the reconstructed images have significantly increased.

Keywords