NVIDIA IndeX for arivis5D Cloud Platform

Key Takeaways

in this demo we'll show how nvidia partnered with arrivas to improve workflows for microscopy researchers nvidia index is a 3d volumetric interactive visualization tool running on nvidia gpus a revis 5d is a web-based image management solution that integrates seamlessly with index very high resolution microscopy can produce up to three terabytes of images per hour processing that data and visualizing the output can quickly become a bottleneck this type of 3d research imaging involves multiple steps to get from the microscope to insights in order to achieve useful results these steps may need to be repeated slowing the journey to discovery the visualization step can be a major obstacle often researchers are either forced to process and render only a selected portion of an image or have to settle for static views or wait for next day 3d renders they may need to re-render multiple times until they find what they're looking for to solve this problem a revis 5d with nvidia index gives the researcher instantaneous visual feedback making image analysis decisions more confidently and gaining insights earlier a revis 5d handles any number of images of any size connecting visualization and image analysis with documents workflows and reports aribus 5d sits on top of nvidia software and hardware stacks this integration can harness anywhere from a single gpu to large clusters enabling volume data display at scale researchers can now work with massive data sets make modifications in real time and more easily locate the most pertinent parts of the data let's take a look at how revis 5d with nvidia index works using a real world example we select an image of the mouse hippocampus to give to index for rendering let's interactively examine this image in full resolution in 3d these cells in blue green and purple make memories in the mouse hippocampus researchers who study how memory can be recovered or improved need to count how many cells there are here we see a cell counting procedure in arrivas but before we can collect this data images have to be analyzed with interactive visualization this analysis can be rapidly accomplished let's create a simple analysis pipeline in the revis vision 4d application we add operations to the panel and interactively set parameters with immediate visual feedback the expected diameter is set and the splitting sensitivity is adjusted this pipeline now finds the nuclei in this image and can be applied to find these objects in other similar data sets as well now that the analysis pipeline is completed it can be uploaded to the rivas 5d server here it is stored and shared enabling collaboration pipeline review and standardization here we see the analysis pipeline running on the server and the segmentation results mapped to the 3d volume we can show the results overlaid on the image data or on their own with this system images are easily explored and analyzed in their entirety while the original data and results are immediately shareable on the web the nvidia index for a revis 5d cloud platform makes it easier to visualize explore and analyze terabytes of multi-channel image volumes this gives researchers a faster path from data to insight you