Denoising is one of the main issues in digital image processing systems. This paper presents a new idea to improve the Peak Signal to Noise Ratio (PSNR) of the denoised image by dividing the image into many blocks then a variable threshold is set according to the energy content of a particular block in the frequency domain obtained using the Contourlet Transform (CT). Computer simulation results show that the proposed method is very efficient to denoise Magnetic Resonance Imaging (MRI) pictures and outperforms the traditional denoising methods that deal with the image as a whole. Experimental evaluation showed that the variable block threshold is more efficient when applied at MRI than other type of images, due to the low energy of the surrounding areas in the MRI.