## MCS-MRI reconstruction on the acquired noiselet encoded and Fourier encoded data for different acceleration factors (up/down: phase encodes, left/right: frequency encode).

**RF**: shows reference image reconstructed from fully sampled Fourier encoded data; **RN**: shows reference image reconstructed from fully sampled Noiselet encoded data; **(a)-(c)**: show images reconstructed using Fourier encoding for acceleration factor of 4, 8 and 16 respectively; **(d)-(f)**: show the difference images using Fourier encoding for acceleration factor of 4, 8 and 16 respectively; **(g)-(i)**: show images reconstructed using noiselet encoding for acceleration factor of 4, 8 and 16 respectively; **(j)-(l)**: show the difference images using noiselet encoding for acceleration factor of 4, 8 and 16 respectively. The result here aligns with the simulation results and noiselet encoding outperforms Fourier encoding in preserving resolution. **(A-H)**: Zoomed portion of phantom images reconstructed with Fourier encoding and noiselet encoding with different acceleration factors. **(A)**: shows the original image reconstructed from fully sampled Fourier encoded data; **(B), (C) and (D)**: show the Fourier encoded reconstructed images for acceleration factors of 4, 8 and 16 respectively; **(E)**: shows the image reconstructed from fully sampled noiselet encoded data; **(F), (G) and (H)**: show the noiselet encoded reconstructed images for acceleration factors of 4, 8 and 16 respectively demonstrating that noiselet encoding produces improved resolution images than than Fourier encoding at all acceleration factors.