> ## Documentation Index > Fetch the complete documentation index at: https://docs.purna.ai/llms.txt > Use this file to discover all available pages before exploring further. # Slides Viewer > Open supported microscopy and whole-slide images, save personal regions, and share selected data with agents. The Slides Viewer lets you inspect supported microscopy and whole-slide images inside your Purna workspace. Open a slide alongside your conversation, zoom into tissue, and save regions with names and notes without changing the original image. ![Purna Slides Viewer for inspecting microscopy images and selecting regions of interest](https://cms.purna.ai/_emdash/api/media/file/01M6JXE409HEW3WBZRF1RT88XS.01M6JXE586QXG9Z1FWJV1JFRRN.webp) Inspect a slide and keep regions of interest alongside your workspace. Opening a slide does not identify cells, segment tissue, calculate statistics, or perform a biological analysis. The controls available depend on the information already present in your file. ## Supported files | Format | File extensions | Requirements | | - | - | - | | Aperio whole-slide image | `.svs` | A complete, readable SVS file. Supported SVS images can include JPEG or JPEG2000 compression. | | Hamamatsu whole-slide image | `.ndpi` | A complete, readable NDPI file. Support depends on the image layout and encoding. | | Prepared OME-TIFF | `.ome.tif`, `.ome.tiff` | One image series, tiled image planes, valid OME metadata, and supported compression. All image data must be in the same file. | | Prepared OME-Zarr archive | `.ome.zarr.zip` | One supported Zarr v2 OME image at the archive root, with image levels, axes, and channel metadata. ZIP entries must use Store mode, without ZIP-level compression. | **The slide size limit is 500 MiB per file.** A successful upload does not guarantee that every image layout or encoding can be viewed. Extensions are case-insensitive; keep the full filename ending, such as `sample.ome.tiff` or `sample.ome.zarr.zip`. Renaming a normal TIFF to `.ome.tiff`, or an arbitrary ZIP to `.ome.zarr.zip`, does not convert its contents. ### Check OME files before uploading * **OME-TIFF:** export a tiled, single-image file. Supported compression includes uncompressed, LZW, Deflate, PackBits, and JPEG. OME-TIFF using JPEG2000 is not supported, even though supported SVS files may use it. * **OME-Zarr:** use Zarr v2 and put the image store directly at the ZIP root. Do not add an enclosing folder or combine several images into one archive. * **ZIP packaging:** choose **Store**, also called `ZIP_STORED` or no compression, for the surrounding archive. Compression inside the Zarr data is separate; not every internal encoding is supported. * **Channels and image levels:** OME images support up to 64 available channels and 32 image levels. You can display up to six channels together. Images must also stay within supported metadata and image-chunk sizes. If your export does not meet these requirements, create a compatible viewing copy in your image-processing tool and keep the original. Purna does not automatically repair, repack, or convert incompatible slides. ## How your file is prepared 1. Upload a supported file through **Files**, or attach it to a conversation. Wait for the upload to finish. 2. Purna checks that the image is readable and prepares its viewing information, including dimensions, available image levels, channels, and physical scale where supplied. 3. Open the file to view it. While preparation is pending, the tab shows **Preparing viewer…** and continues automatically when ready. New supported uploads prepare automatically. An older file may need preparation when you first open it. Closing the tab does not cancel preparation, and reopening a ready file reuses its prepared viewing information. Your original remains one visible file in **Files**. Preparation does not overwrite image pixels, change coordinates, infer cell types, detect tissue boundaries, or calculate an analysis. It does not create missing calibration or turn an ordinary image into a compatible slide. Large files can take longer to prepare and open. The viewer loads the image detail needed for your current view as you navigate. Changing zoom or moving to another region can load more detail; a first visit may take longer than revisiting the same area. If preparation fails, the original remains available for download where you still have access. An agent can help investigate the file or write a script for a compatible viewing copy as a separate task. Review any conversion and its assumptions yourself. ## Working with your data ### Open and arrange a slide Open the file from **Files** or a chat attachment. On desktop, drag the workspace divider to resize the viewer pane. Use the expand icon for a larger canvas and collapse it to return to your workspace. Use the toolbar to pan, zoom in or out, and fit the image to the available space. Where offered, rotate the slide to inspect it from another orientation. Download the original with the download icon. You can switch between file tabs while working. The viewer retains your current view across tab switches. Wait for **Saved** before closing or reloading if you have edited a region or its notes. ### Adjust image display For supported OME images, open the available display controls to choose channels, adjust their colors and intensity ranges, or switch Z and T planes when those dimensions exist. RGB images use their original color mapping. These settings change the display, not the original measurements. A brighter channel or different color does not establish a biological difference. Keep the selected plane and channel settings in mind when comparing regions. ### Select a region 1. Choose the rectangle or polygon tool. 2. Draw around the area you want to inspect. 3. Open **Regions** and select the region if its details are not already visible. 4. Enter a meaningful name and optional notes, such as `Tissue edge — check staining`. Regions refer to positions in the original image. Zooming, rotating, or resizing the viewer does not move the saved region within the source image. For images with multiple planes, a region also retains its selected plane and display context. The viewer shows region dimensions in pixels. Physical measurements appear only when the source contains valid pixel calibration. Do not infer micrometers from an uncalibrated image. ### Save and revisit regions Regions save automatically. Use the save icon to save the current edit immediately, and wait for **Saved** before leaving. The delete icon removes the selected personal region; it does not delete or alter the image. Your named regions, notes, viewport, and supported display settings belong to your personal view of that file. They are not edits to the source or shared annotation editing. Reopen the same file to revisit successfully saved regions. Replacing the source file starts a new view. Regions from the previous file are not automatically transferred to its replacement. Downloading the original does not include your personal regions and notes. ### Select cells when your data supports it A slide region is an area of an image, not a list of identified cells. Drawing around tissue does not detect, count, label, or segment cells automatically. For prepared datasets with cell identities and coordinates, use the [Spatial Omics Viewer](https://docs.purna.ai/scientific-components/spatial-omics-viewer.md). Its available tools can select cells with a lasso, select existing cell groups, and coordinate highlighting across linked views. Those capabilities require prepared cell data; a slide image alone does not supply it. ### Use a region in chat 1. Select a region and review its name, notes, image plane, and display settings. 2. Choose **Use in Chat**. 3. Choose **Start a new chat**, or an existing chat in the image's project. The dialog lists recent chats; type to filter them. 4. Review the region preview, region-details file, and source-file reference added to the draft. 5. Add your question and send the message when ready. Attaching a region does not send a message or start an analysis. Only the selected region's details are shared; unrelated personal regions are not automatically included. The shared region is a snapshot. If you edit it later, attach a new snapshot when you want the chat to use the revised selection. The source reference identifies the full slide. The selected region covers only a smaller part of it. A region preview is for visual reference and is not a substitute for original values in quantitative work. ## Working with agents Agents can work with your files and read the selection details you explicitly share, including region coordinates, names, and notes. Use them to help with analysis, statistics, and scripts. They can help check metadata, describe a reproducible analysis, or write code to examine original data. For example, you can ask: * "Read the selected region's coordinates and calibration. Write a script to report its dimensions, and explain any missing information." * "Help me design a comparison between this region and surrounding tissue. List the required measurements and statistical assumptions before calculating anything." * "Check whether this slide contains the channels and metadata needed for my analysis. Do not infer cell identities from the region name." > [!WARNING] > > **Agents cannot view or operate the live graphics or slides in your viewer.** They do not share your screen or continuously see your navigation and selections. Sharing selection details or a region preview does not mean an agent has visually reviewed the full slide. > > Purna is designed to keep a **human in the loop**. Inspect the slide yourself and verify the selection, source file, measurements, assumptions, statistics, and scripts. **Do not rely completely on the agent** or treat its response as a validated biological finding or diagnosis. An agent's response does not automatically change your image or add regions to the viewer. Give explicit instructions for any separate analysis or conversion you want performed, and review its outputs. ## Limitations and files that will not work | Limit or unsupported input | What to do | | - | - | | Slides larger than 500 MiB | Create a smaller compatible viewing copy or use a suitable external viewer. Multi-gigabyte slides are outside the current limit. | | Ordinary `.tif` or `.tiff`, PNG, JPEG, or other images renamed as slides | Use the appropriate normal preview, or export a genuine supported slide format. | | DICOM, CZI, MRXS, SCN, BIF, and other unlisted slide formats | These are not native Slides Viewer inputs in this release. Use an external viewer or explicitly prepare a supported copy. | | Untiled OME-TIFF, multiple image series, or images that rely on separate TIFF files | Export one self-contained, tiled OME image. | | OME-TIFF using JPEG2000 or another unsupported encoding | Export with a supported compression setting. | | Zarr folders, Zarr v3, nested OME stores, or archives containing several images | Export a single supported Zarr v2 OME image at the root of an `.ome.zarr.zip` archive. | | ZIP-level compression, damaged archives, or incomplete uploads | Re-export with ZIP Store mode and verify the archive opens locally before uploading. | | Images with oversized chunks, excessive metadata, or unsupported layouts | Prepare a simpler viewing copy. File size alone does not establish compatibility. | | More than 128 saved regions, or polygons with more than 1,024 vertices | Remove unused regions or simplify the polygon. A saved view also has a 2 MiB total limit, including notes. | | Missing calibration, cell identities, or biological annotations | The viewer does not invent this information. Use pixel dimensions or prepare the required data separately. | The interactive Slides Viewer is a desktop web feature. Mobile viewing does not provide the same interactive tools. Large images, many channels, and limited browser memory can affect responsiveness even when a file is supported. Automatic segmentation, cell detection, biological interpretation, shared annotation editing, and agent-created viewer overlays are not part of this release. ## Troubleshooting | What you see | What to try | | - | - | | **Preparing viewer…** stays visible | Confirm the upload finished. Allow more time for a large file. Use **Retry** if offered or reopen the file. If it remains stuck, contact support with the filename and exact error rather than repeatedly uploading duplicates. | | Unsupported format, series, compression, or image layout | Compare the export settings with the supported-file requirements above. A filename change cannot repair the contents. Export a compatible viewing copy. | | "Could not find end of central directory" or another ZIP error | Check that the archive is complete and opens locally. Confirm it contains one OME-Zarr image at the root and uses ZIP Store mode. | | A blank or unexpectedly dark OME image | Check the selected channels, colors, intensity ranges, and Z/T plane. Ensure the chosen channels actually contain signal. | | Loading continues while zooming or panning | Allow the view to settle while the required image detail loads. If loading repeatedly continues while untouched, use **Retry** or reopen the tab and report it with the filename and error. | | A region is missing after changing planes | Return to the region's original Z/T plane and select it from **Regions**. Regions retain their image-plane context. | | **Not saved**, a save failure, or "saved view exceeds its limits" | Keep the tab open. Check your connection, remove unused regions, shorten long notes, or simplify polygons. Use the save icon and wait for **Saved**. | | "This view changed in another tab" | Preserve any unsaved notes you need, then choose **Reload saved view**. Another tab saved a newer view; edits are not automatically merged. | | The file changed or old regions are unavailable | Close and reopen the current file. Verify regions against the new image; annotations are not automatically transferred to a replacement. | | **Use in Chat** is unavailable or fails | Select a region, confirm access to the image's project, and choose a chat in that project. Retry after resolving the visible error. | | Physical measurements are missing | The image may lack valid calibration. Use pixel dimensions and check the original export metadata before reporting a physical size. | | Access denied or unavailable image | Confirm that you still have access to the file and its project. Reopen after access is restored. | | Blank canvas, graphics error, or slow interaction | Check browser graphics support and available memory. Try fewer displayed channels or a smaller compatible viewing copy. | When contacting support, include the filename, extension, file size, exact error, and the action that failed. Avoid sharing sensitive image contents unless requested through an appropriate support channel. Return to [Scientific Components](https://docs.purna.ai/scientific-components/overview.md), explore cell-based data in the [Spatial Omics Viewer](https://docs.purna.ai/scientific-components/spatial-omics-viewer.md), or see [Files](https://docs.purna.ai/guides/files.md) for uploading and organizing your data. This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.