In a wide range of scientific fields, 3D datasets production capabilities have widely evolved in recent years, especially with the rapid increase in their sizes. As a result, many large-scale applications, including visualization or processing, have become challenging to address. A solution to this issue lies in providing out-of-core algorithms specifically designed to handle datasets significantly larger than memory. In this article, we present a new approach that extends the broad interactive addressing principles already established in the field of out-of-core volume rendering on GPUs to allow on-demand processing during the visualization stage. We propose a pipeline designed to manage data as regular 3D grids regardless of the underlying application. It relies on a caching approach with a virtual memory addressing system coupled to an efficient parallel management on GPU to provide efficient access to data in interactive time. It allows any visualization or processing application to leverage the flexibility of its structure by managing multi-modality datasets. Furthermore, we show that our system delivers good performance on a single standard PC with low memory budget on the GPU.

Large scale images allow pathologists to perform reviews, using computer workstations instead of microscopes. This trend raises a wide range of issues related to the management of these massive datasets. In particular, efficient solutions for data storage and processing have to be developed in order to deliver increasingly reliable and faster analyses. In addition, the improvement of workflows also requires the reinforcement of visualization capabilities. In this paper, we present a new virtual microscopy (VM) approach for interactive-time navigation in large images stack with 3D visualization on multi-stereo display. Our work is based on a tool that combines 3D pyramidal representation and out-of-core data management for interactive on-demand streaming of large datasets on GPUs. The preliminary results suggest that the proposed solution facilitates the reading and the understanding of data essentially because they are spatialized.