Interactive Visualization of Multi-Terabyte Datasets on Commodity Hardware

Cultured cells have long been used to study cellular response to drug treatments as surro- gates to how cells would respond in living organisms. Traditionally, 2D cell cultures have been used for these assays, but 3D cultures (for e.g. organoids) are rapidly growing as a better analog to the cellular environment in our own bodies. Many of these assays are based on observing the cellular response with 3D microscopes such as confocal and lightsheet. These experiments are based on terabytes of image data, thousands of separate multi-dimensional image objects, and millions of identified sub-image ob- jects. They require a software system for managing, visualizing and analyzing these large amounts of 3D image data and associated meta-data. Here we propose to commercialize our cloud-based system for visualization and AI-based analysis of large 3D image data sets to address these needs in studies using organoids and whole-brain imaging. In Phase I and Phase II, we developed a cloud-based platform for storing, searching and analyzing large multi-dimensional image-like datasets with the idea that large data should live in one scalable central location with the analysis tools as close as possible to the data, and the visualization tools accessing it remotely using commodity hardware and software. In Phase I and II, we have developed this platform and shown its utility in several areas where high-end imaging, visualization or image processing is needed, such as materials processing, environmental monitoring, remote sensing and life sciences. In parallel to this project, we have also developed easy to use 2D AI tools, which we intend to apply to these 3D datasets. For Phase IIB, we have identified the growing need for 3D imaging as a commercial application in the life sciences as evidenced by the current growth in the use of image-based assays for 3D cultures, and lightsheet microscopy of whole-tissue samples, such as whole-brain imaging. Three dimensional cell cultures are being increasingly used by phar- maceuticals to study potential new drugs. We intend to market this system for drug-discovery in 3D cultures, including high-end 3D rendering viewers deployed in web browsers, and AI-based analysis sys- tems designed for ease of use by bench scientists. Our 3D capabilities are unique and address pain-points in this emerging market. Our system is also capable of addressing traditional 2D imaging, but here the pain-points have been partially met with desktop or home-made solutions which are much harder to solve for 3D. Essentially, we are set to tackle the scientific imaging market as a whole, by initially focusing on 3D, where we can offer a unique solution.