Time to take a look at another any-widget based Curvenote extension, one that we can use to visualize huge (or small) microscopy datasets in articles through client-side javascript libraries that point to Zarr datasets server from standard cloud object storage (S3 and the like).

Figure 1 is just such a visualization of a 1.3TB image of a zebrafish embryo, with tiles loaded on demand as a reader pans and zooms around to explore the data.

More Examples¶

• Multicolor multiscale brain imaging

• SARS-CoV-2 infected intestine organoids

• Plant germline differentiation within Arabidopsis flowers

The Viz Library & Wrapper¶

vizarr is a minimal, client-side image viewer developed by Harvard Medical School’s Department of Biomedical Informatics (HMS-DBMI) for visualizing Zarr-based images, particularly those following the OME-NGFF standard. Built on top of the Viv library for GPU-accelerated rendering, vizarr enables efficient, interactive exploration of large, multiscale microscopy datasets directly in web browsers without requiring server-side processing. The viewer was originally developed to support interactive multimodal image alignment workflows, but it’s existence and ease of use has seen it become more widely.

Using the any-widget plugin interface in Curvenote, vizarr can be embedded directly in any article allowing authors to include data directely in interactive figures 🚀.

How to use it¶

The image above is rendered using the following MyST Markdown directive:

:::{any:widget} https://curvenote.github.io/widgets/widgets/vizarr-viewer.js
:class: w-full

{
    "source": "https://uk1s3.embassy.ebi.ac.uk/idr/zarr/v0.1/4495402.zarr",
    "height": "600px"
}

:::

Where we point out to a stable-hosted version of widget at https://curvenote.github.io/widgets/widgets/vizarr-viewer.js and include a URL to our dataset in the source field of the directive body.

The Dataset¶

If you regularly use microscopy data, likely you already know OME-Zarr, if not the you may still be able to leverage the viewer provided you prepare your dataset accordingly. Here’s some background info the might help.

What is OME?¶

OME^[1] addresses a critical challenge in scientific imaging: the proliferation of proprietary file formats from different microscope manufacturers makes it difficult to share, analyze, and preserve microscopy data across institutions and over time. By providing standardized, open formats and tools, OME enables researchers to work with imaging data regardless of its origin, facilitating collaboration and reproducibility in the life sciences.

What is OME-Zarr¶

OME-Zarr is a next-generation file format specification that builds on the Zarr^[2] chunked array format to store large multi-dimensional microscopy images in a cloud-optimized, compressed format. OME-Zarr breaks images into smaller chunks that can be efficiently accessed, transferred, and processed in parallel. This makes it ideal for large datasets, cloud-based analysis, and interactive visualization—enabling researchers to work with terabyte-scale imaging data without downloading entire files.

OME-Zarr vs OME-Tiff¶

Unlike traditional formats like OME-TIFF, which store images as single files, OME-Zarr breaks images into smaller chunks that can be efficiently accessed, transferred, and processed in parallel. This makes it ideal for large datasets, cloud-based analysis, and interactive visualization—enabling researchers to work with terabyte-scale imaging data without downloading entire files.

The format is part of the broader OME-NGFF (Next Generation File Format) initiative, which aims to modernize how scientific imaging data is stored and accessed.