Frequently Asked Questions

Zeta Imaging is a research initiative based at the University of British Columbia (UBC) focused on developing high-performance software tools for basic science research, with a particular emphasis on microscopy imaging.

ZetaCore is: a modular, high-performance processing pipeline for lattice light-sheet fluorescence microscopy (LLS). It's designed to process terabyte-scale datasets efficiently using GPU acceleration while maintaining a constant memory footprint.

ZetaCore is currently in alpha stage. We're working hard to prepare it for public release. Sign up for our mailing list to be notified when it becomes available and to receive early access opportunities.

Our benchmarks show that ZetaCore is 5-6 times faster than existing solutions. For example, a dataset that would typically take approximately 3 days and 7 hours to process completes in under 16 hours with ZetaCore.

ZetaCore uses a GPU-accelerated Richardson-Lucy (RL) deconvolution algorithm. This is a standard, well-established iterative deconvolution method that's widely used in microscopy. Our implementation is optimized for modern GPUs, allowing for faster processing compared to CPU-based implementations.

The hybrid in-memory/out-of-core approach manages data transfer between system RAM and GPU VRAM to maximize performance while maintaining a constant memory footprint.

ZetaCore uses a global flat-field correction approach that specifically targets illumination artifacts caused by the microscope hardware while preserving sample-specific emission patterns. The algorithm creates a correction image by preserving low spatial frequency information.

ZetaCore implements a fully out-of-core GPU-accelerated deskewing engine. This means that even datasets much larger than your available GPU memory can be deskewed efficiently. The engine uses memory-mapped files and intelligent chunking to process the data in manageable pieces while maintaining GPU acceleration throughout.

Minimum: NVIDIA GPU with 8GB VRAM, 32GB RAM, plenty of SSD storage, Windows 10/11 or Linux.

Recommended: NVIDIA GPU with 16GB+ VRAM, 64GB+ RAM, we highly recommend NVMe SSD storage for optimal performance.

Yes! That's one of ZetaCore's key innovations. Our out-of-core processing techniques and memory-mapped file system allow you to process datasets that exceed your available GPU VRAM or system RAM while maintaining a constant memory footprint.

ZetaCore currently has native support for Zeiss CZI files and TIFF files. If you have a file format that is not supported, please let us know.

ZetaCore includes global flat-field correction, Richardson-Lucy deconvolution, deskewing, and an efficient stitching engine. All core processing stages are fully GPU-accelerated.

We'll be launching a beta program soon! Sign up for our mailing list and indicate your interest in sharing imaging data with us. We'll prioritize beta invitations for users who can help us test with diverse datasets.

The best way to reach us is through the signup form. Include your questions or comments in the optional comments field, and we'll get back to you as soon as possible.

Yes! We're always interested in collaborating with researchers and testing ZetaCore on diverse datasets. If you'd like to share imaging data or collaborate, please check the appropriate box on our signup form.