1: \begin{abstract}

2: The Fast Proximal Gradient Method (FPGM) and the Monotone FPGM (MFPGM)

3: for minimization of nonsmooth convex functions are introduced and

4: applied to tomographic image reconstruction. Convergence properties

5: of the sequence of objective function values are derived, including

6: a $O\left(1/k^{2}\right)$ non-asymptotic bound. The presented theory

7: broadens current knowledge and explains the convergence behavior of

8: certain methods that are known to present good practical performance.

9: Numerical experimentation involving computerized tomography image

10: reconstruction shows the methods to be competitive in practical scenarios.

11: Experimental comparison with Algebraic Reconstruction Techniques are

12: performed uncovering certain behaviors of accelerated Proximal Gradient

13: algorithms that apparently have not yet been noticed when these are

14: applied to tomographic image reconstruction.\\

15: \textbf{Keywords}$\;$Computerized tomography imaging, Convex optimization,

16: Proximal gradient methods, Iterative algorithms.\\

17: \textbf{Mathematics Subject Classificatio}n$\;$65K05, 68U10, 90C25,

18: 90C06, 92C55, 94A08

19: \end{abstract}

20: