Dynamic metasurface apartures (DMAs) enable the use of computational imaging techniques, which offer significant advantages in terms of hardware with respect to traditional synthetic aperture radar (SAR) systems. However, since the signal on the receiver side is usually compressed to one or a few channels, the signal processing layer becomes more complicated. Traditionally, spatial domain image reconstruction algorithms have been proposed to address this limitation, which involve the computation of a usually ill-posed sensing matrix. As a result, spatial domain techniques can be computationally expensive and time consuming. Alternatively, recent studies have proven that a pre-processing decompression step can be employed to decompress the signal and transform it into the form of a conventional multistatic signal. Then, image reconstruction can be performed fully in the frequency domain using range migration techniques. In this paper, we develop a decompression step for a stationary multistatic DMA-based system used for three-dimensional (3D) near-field imaging. After the proposed step is applied, the image can be reconstructed using the conventional range migration algorithm (RMA). The simulation results verify that high-quality images can be produced with the proposed technique. Additionally, the overall reconstruction time is significantly reduced in comparison with traditional spatial domain algorithms.