Advancing Substructure Analysis in Tomograms

Identifying and isolating specific parts within three-dimensional (3D) tomograms present a significant challenge. This study leverages computational techniques and data analysis tools to dissect these dense objects virtually. In particular, this study is concerned with understanding how clustering algorithms such as k-means and, eventually, more sophisticated and complex methodologies can be used for this segmentation task. Indeed, although other more complex algorithms already exist, they need to work on specific tomograms, such as single cells. In contrast, a more general solution is required when those constraints cannot be guaranteed. The methodologies pave the way for further advancements in understanding complex systems and have potential in industrial applications. Moreover, it suggests that such tools can be integrated into interactive systems for immersive and efficient results visualization and analysis.