Creating a volumetric 3D scan of an object using CT, MRI , or other imaging modalities, is currently a sequential process. First a scan is made using the physical scanner, then the 3D image is computed based on the acquired dataset, and afterwards the resulting 3D volume is analyzed by an expert on the specific questions that one needs to answer about the object.
In many cases, however, the specific questions and the data that is needed to answer those questions are not well-defined before the scan. At the Computational Imaging group of CWI , we are developing a broad spectrum of techniques and algorithms aimed at redesigning the 3D scanning process into an interactive process, where the reconstructed image is computed online during the scan, the user (or data-driven algorithms) can analyze the data, and the scanning process can be steered dynamically to explore the object in real-time.
In this lecture I will sketch this general vision of “explorative 3D imaging” and discuss several of the technical, mathematical and algorithmic approaches involved in creating a fully interactive 3D imaging pipeline.