Diffusion-weighted imaging (DWI) is an essential sequence in many clinical MRI protocols but remains time-consuming and hardware-intensive, particularly in mid- and low-field systems. This study proposes a method for generating synthetic DWI (DWI-Syn) using Intravoxel Incoherent Motion (IVIM) modeling to reduce acquisition time while maintaining diagnostic quality. The study utilized two datasets: 10 prostate DWI acquired at 1.5 T with eight b-values (0–1800 s/mm2) and 10 brain DWI acquired at 0.35 T with four b-values (0–500 s/mm2). A custom Python-based pipeline was developed to correct for both patient and internal organ motion using BRISK, ORB, and SIFT feature-based registration algorithms. The optimal transformation was selected using the SSIM correlation coefficient. Segmentation was automated using a U-Net model trained on manually annotated masks created in 3D Slicer. Voxel-wise IVIM fitting was performed to estimate diffusion (D), pseudodiffusion (D∗), and perfusion fraction (f), which were used to generate DWI-Syn at additional b-values. DWI-Syn demonstrated strong correlation with original data using SSIM above 0.75, improved contrast homogeneity, and reduced noise, enhancing tissue delineation in the prostate and extending functional b-value coverage in low-field brain imaging. Limitations were noted at b-values above 1000 s/mm2 in low-field settings due to low signal-to-noise ratio. The U-Net model achieved a final validation Dice Similarity Coefficient of 0.9285, with a validation loss of 0.0088. The proposed method supports the feasibility of IVIM-based DWI-Syn as a strategy to improve imaging efficiency and clinical flexibility in resource-constrained MRI systems. Further validation is warranted to optimize b-value protocols and support broader clinical integration. The limited number of patients included in this study constrains the generalizability of the findings and highlights the need for further validation on larger cohorts.