Abstract. Computed tomography is a powerful tool for plasma diagnostics. Unfortunately, viewing access in many plasma experiments is strongly limited, which leads to a highly underdetermined ill-posed problem. A maximum a posteriori (MAP) algorithm is a promising statistical approach to this problem. In this paper, we study the MAP algorithm which yields the most probable image estimate from limited and noisy data. An adaptive chi-square statistical criterion is proposed to control the quality of reconstruction. An improvement in reconstruction image quality over conventional tomographic methods, such as the algebraic reconstruction technique and maximum entropy algorithm, is illustrated by several numerical examples.

Print publication: Issue 3 (August 2004)
Received 7 July 2004
Published 5 August 2004