Add support for CG method for quick reconstruction of the images

[chaithyagr]

Currently, we support all the ingridients needed for doing CG iteration, just a small function in extras module with:

Here is a simple example of gradient descent step. All thats needed is to add momentum required.

def cg(fourier_op, kspace_data, num_iterations):
    L = fourier_op.get_lipschitz_cst()
    image = np.zeros(fourier_op.shape)
    for i in range(num_iterations):
         grad = fourier_op.data_consistency(image, kspace_data)
         image = image - grad * L
    return image