Diagnostic Viewer

Cloud viewer software
that speeds you up

The first diagnostic viewer built to deliver the full power of the cloud.

Your viewer software was built for an on-prem world.

As a result, hosting it in the cloud makes it slower and more expensive.

This creates practice-wide friction to cloud adoption that makes it impossible to modernize.

How Sirona is Different

Edge Streaming

Sirona’s viewer understands where each radiologist is, and uses infrastructure designed by Netflix to cache data closer to you and stream data over a proprietary world-wide network.

Viewer performance

Load images reliably from anywhere on earth

Pre-Cached: Instant

(Coming Soon!)

Hot-Cached: Industry Standard

(Coming Soon!)

Cold-Cached: Industry Leading

(Coming Soon!)

Platform AI

Save time every day with
powerful, embedded AI

Hanging protocols that work

Set them up once and let Sirona’s series and study classification engine do the rest.

AI hanging protocols

Spines that label themselves

Never manually label disc-levels for spine MRI's again.

Auto-measure lung nodules

Click on a lung node, and measure it.

Embed 3rd party AI

Sirona is built to orchestrate 3rd party AI products.

Unified Data Model

Link pixels to words
at every level

Sirona ties all your data together — images, reports, speech, AI findings — into a Unified Data model powered by a proprietary ontology, SironaLex. This facilitates a real-time, bidirectional connection between pixels and words that is the crucial building-block for agentic AI workflows.

From studies to reports

Automatically populate the priors section of reports for every prior you hang.

From segmentations to sections

Automatically navigate images by selecting their corresponding sections in reports.

From images to findings

Navigate to key findings from the report text. Linkage can be AI or manually generated.

From measurements to words

Make a measurement, and insert the number into the report automatically.

Speech-to-Action

Turn speech into automation

Voice Commands

Activate any tool and action platform-wide

Activate any tool and action platform-wide using voice commands. Example: “Switch to length, set window level to lung.”

Learn more

Clinical Queries (Coming Soon)

Ask conditional questions of the clinical record

Ask conditional questions of the clinical record, and have action taken based on the answers. Example: “If priors mention liver nodules hang the most recent study that mentions them, show me the biggest nodule, and switch to lung tool.”

Learn more

Agentic Automations (Coming Soon)

Combine questions, commands, and findings

Combine questions, commands, and findings into a single stream of dictation. Speech-to-insight. Speech-to-action. Combined: Speech-to-automation.

Learn more

Subspecialty Viewing

Ditch PET & Mammo Viewers

Sirona is focused on expanding our mammography and PET viewing capabilities so practices can eliminate expensive specialty viewing apps.

PET Viewing (FDA Cleared)

Sirona supports PET viewing, including PET/CT fusion, MIP's, SUV measurements, and more natively.

Mammo Viewing

Sirona supports mammography viewing, and is expanding the platform to support a broader suite of mammo-specific tooling this year.

Built Different

View your images anywhere

Enabled by the RadOS platform that powers it, Sirona’s viewer redefines accessibility in radiology viewing. Read anything, anywhere, with any internet that meets our basic requirements.

FAQs

What internet performance does your viewer require?

>100ms per second is optimal. If your local ISP doesn't offer this, Sirona is compatible with Starlink.

What workstation compute requirements do you recommend?

Any modern workstation with a current browser version and 16GB+ RAM is sufficient. Sirona runs browser-based with server-side image processing, so workstation requirements are minimal relative to legacy on-prem viewers. For mammography workstations, we additionally require a 5MP+ display at 60Hz.

How many monitors do you support? Can you support multiple 5-6MP displays?

We do support multi-monitor, up to 4 distinct monitors, at a time. These can be placed in any combination of landscape or portrait. For monitor configurations: • 2MP minimum resolution. • We require a minimum of 5MP display resolution required for mammography viewing and do support 6MP. • 60 Hz refresh rate minimum.

Does your platform support local caching or intelligent pre-fetching of image data to minimize perceived latency, or just Edge Streaming?

Our browser-based caching is currently supported. We are exploring various caching strategies that take increasingly into account client behavior and available bandwidth to further optimize performance.

Does your viewer architecture leverage discrete GPUs effectively (Ex. NVIDIA RTX series cards)?

Yes, the viewer architecture leverages these discrete GPUs effectively. Our architecture is based on VTK.js (which uses WebGL 2.0 today, with work in progress for WebGPU). The calls we need for rendering image data do eventually route through ANGLE, but are appropriately mapped to the proper discrete GPU commands and run on GPU, not the CPU or integrated GPU (as long as the device is not on a deny-list). We verify this by checking with Chrome’s telemetry (“chrome://gpu”) to ensure that WebGL is marked as “Hardware accelerated” and has the appropriate GPU marked as the active GPU.

Is it possible to observe in-session GPU utilization?

Network related timings can be observed through the Chrome developer tools. GPU utilization can be observed via profiling. Correct GPU assignment can be viewed in “chrome://gpu”.

What steps have you taken to optimize GPU and system resource usage for high-demand workflows — such as breast imaging, multi-planar reformatting, or cine loops?

Sirona’s performance strategy emphasizes server-side processing — including frame extraction, compression and browser caching — rather than pushing computational load (like pixel decoding) to the browser. This architecture inherently minimizes reliance on local GPU capabilities, and instead focuses on fast, efficient delivery of already processed and compressed frames to the viewer.

Do Chrome-imposed memory constraints on browser tabs affect performance?

While older versions of chrome could be limited to 2-4GB depending on if you ran 32-64bit operating systems, new chrome instances running on Windows OS are capable of supporting 16GB per tab, providing sufficient space to run over viewer application.

Is there scrolling lag in Sirona’s viewer like there is with other browser-based viewers, and if not, why?

No, there’s no lag. Once image data is fully loaded into the browser, there is no lag when scrolling through images. This lag is typically a function of viewers that use “server side rendering; Sirona’s Edge Streaming engine gets data to the browser faster and more reliably, but the rendering engine itself is client-side, making it the best of both worlds. When scrolling through images prior to having them in browser memory, we adjust the prioritization of image downloading appropriately to target the desired frame and those around it.

Do you support 10-bit grayscale or DICOM GSDF calibration natively? How do you handle those requirements for mammography?

While the default browser bit depth is 8-pixels, Chrome does offer support for higher bit depth with proper settings enabled and a monitor which supports 10-bit display.