In this presentation, an overview of the liver segmentation problem, its importance and challenges will be introduced first. A brief survey of different segmentation approaches found in literature is then presented.

After that, I will explain our novel shape-based liver segmentation approach. This approach, unlike the other shape-based approaches, is model-free, and it does not require prior shape or intensity models construction. In contrary, it exploit the relation between consequent slices in multi-slices CT images to estimate and propagate shape and intensity constrains. Then, these constrains are integrated with a shape-based graph cuts algorithm to extract the liver object in each slice.

This approach needs a simple user interaction and it eliminates the burdens associated with model construction like data collection, manual segmentation, registration, and landmark correspondence. Moreover, it is talented to deal with complex shape and intensity variations. This model free approach was evaluated on 50 CT images from three different datasets with several liver abnormalities, including tumors and cysts, and it was achieved high average gauged scores of 80.4, 79.2, and 81.7 for these datasets. Finally, the preliminary results of tumor detection based on this liver segmentation approach will be shown.