[en] Computed tomography (CT) is a powerful imaging technique, using gamma ray (or X-ray) that can provide a 2D or 3D cross sectional view of the interior of an object as if it had been sliced open along the image plane for viewing. The gamma ray transmission data are collected by an array of detectors at many different angles and these data are then transformed into meaningful cross-sectional image, a reconstruction of the objects interior. The invention of the CT scanner revolutionised the field of medical diagnostic imaging because it provided more detailed and useful information than any previous non-invasive imaging techniques. For the same reason, the method is being used increasingly in process industries for visualising flow in multiphase reactors. One key area of research addresses the selection of appropriate image reconstruction algorithms for evaluating and visualising the process in question. Algebraic algorithms such as Fourier/ convolution techniques when applied to determine the phase holdup distribution in two phase systems in industrial process investigation, either assume the systems to be azimuthally symmetric in distribution or consider the gamma ray transmission process to be deterministic there by completely ignoring the stochastic nature of data. Expectation-maximization (EM) algorithm account for the stochastic nature of the gamma ray transmission across the domain of interest. This makes this algorithm more favorable for image reconstruction to determine the phase holdup distribution. In this work, we implement the EM algorithm for image reconstruction in gamma ray tomography for industrial applications. Experimental gamma ray transmission data obtained with a fan beam geometry gamma ray scanner, and simulated transmission data based on a synthetic phantom, with two phases (water and air) were considered in this study using MATLAB programming. (author)