ScanContext
Free · Local-first · Windows, macOS & Linux

Understand your medical scans on your own computer.

ScanContext is a free app that opens your medical images and reports, organizes them, translates and summarizes them in plain language, and helps you prepare clear questions for your doctor — all without uploading anything.

⬇ Install ScanContext See how it works

Detecting your system…

The ScanContext dashboard — import status, upload, AI provider, and recent studies.
ScanContext in action — your studies, viewer, and reports together in one window.
⚠️

Not a medical device. Not a diagnosis. ScanContext helps you organize, view, and understand medical imaging and reports. It does not diagnose any condition. Every result — especially anything AI-suggested — must be confirmed by a qualified clinician.

What it is

A calm, private place for confusing medical files

A CD of scan images, a folder of DICOM files, a PDF report in a language you don't read — medical records arrive in formats made for hospitals, not for you. ScanContext opens all of them in one window, keeps them on your machine, and explains them in plain language.

🔒

100% local-first

Runs on your own computer. Your scans and reports never leave it unless you deliberately turn on a cloud AI provider.

🩹

For patients & families

No medical or technical background needed. Plain-language summaries, guided tour, and a help button on every screen.

🧾

AI is optional

ScanContext works fully without AI. When you choose to enable it, results are always labelled "candidate findings, not a diagnosis".

How it works

From a pile of files to a clear conversation

Five simple steps. You stay in control at every one.

Import your files

Drag a file or a whole folder straight onto the app — DICOM studies, image files, ZIPs, PDFs, or text reports. ScanContext sorts them into studies for you, and you can delete any study when you're done.

View & annotate

Open images and DICOM scans in a built-in viewer. Draw, mark, and save annotations to remember what you want to ask about.

Extract & translate

Pull the text out of reports and, when needed, translate non-English reports into English — right on your machine.

AI assist (optional)

If you turn it on, AI summarizes reports and points to regions of an image worth a closer look — always clearly marked as candidates, never a diagnosis.

Export for your doctor

Generate a tidy, doctor-friendly summary of the study and your questions to bring to your appointment.

Under the hood: ScanContext runs as a small set of services inside Docker on your computer — a viewer, a report engine, and a local DICOM store. You never touch any of that; the app starts and stops it for you.

See it in action

A look inside ScanContext

Everything below is the real app, captured from its built-in demo study — a fully fictional CT chest scan with invented patient details. No real medical data is shown.

ScanContext dashboard: drag-and-drop upload area, studies, and AI provider status.
Drag a file or a whole folder to import; your studies, import status, and AI provider at a glance. Delete a study anytime.
The ScanContext dashboard in dark theme.
Light, Dark, or System theme — switch from the header or Settings.
The ScanContext image viewer showing a CT slice.
Scroll slices, zoom, and adjust window/level in the viewer.
A CT slice with an annotation box overlaid.
Confirmed marks in solid blue; AI suggestions in dotted amber.
The report panel showing an imported radiology report.
Reports shown as imported, with optional translation and summary.
The AI review tab with the Analyze study action.
Optional AI review — candidate findings, never a diagnosis.
The export tab with the Anonymize option enabled.
Export a doctor-friendly ZIP — anonymized if you choose.

Why use it

The benefits, in plain terms

🏠

Your data stays home

Nothing is uploaded by default. Files live in a folder on your computer that you can see and back up.

💬

Plain language

Turns dense radiology wording into something you can actually read, and translates foreign-language reports.

🧰

One viewer for everything

DICOM, JPEGs, PDFs, ZIP archives, text reports — all opened in one place, no extra software.

💰

Free to use

Built by one independent developer. Free, with an optional small donation if it helps you.

💻

Works everywhere

Windows, macOS, and Linux — the same app, the same simple steps on every system.

🪥

You decide on AI

AI is off until you switch it on, and you confirm cloud use for each individual study.

📋

Ready for the doctor

Walk into your appointment with an organized summary and a clear list of questions.

📱

One-click app

A desktop app with a tray / menu-bar icon starts and stops ScanContext for you — no scripts to run.

⚡

Fast after first run

The first start downloads the app once. Every start after that is quick.

Install — a friendly step-by-step guide

Get ScanContext running in three steps

No GitHub account and no source code needed, and we'll walk you through every step. On Windows and macOS you don't touch a terminal at all. Please follow the steps in order — just pick your system in Step 2.

Step 1 — one time

Install Docker Desktop (free)

ScanContext runs inside Docker, so install Docker Desktop first. It is free and you only ever do this once.

🐋 Get Docker Desktop ↗

Then open Docker Desktop once and wait until it says "running". Leave it running while you use ScanContext. That's the whole of Step 1.
Step 2 — pick your system

Download & start ScanContext latest release

There are two ways to run it — the app (easiest) or a download bundle. Choose whichever you prefer; both work the same once open. The buttons below download the file directly — there's no other page to visit.

Option A — recommended

The ScanContext app

A small desktop app that starts and stops ScanContext for you.

⬇ Download for Windows — installer (.exe) ⬇ .msi alternative
  1. Double-click the downloaded installer and follow it through.
  2. If a blue SmartScreen warning appears, click More info → Run anyway (only the first time — the app isn't code-signed yet).
  3. Open ScanContext from the Start menu. It checks Docker, downloads the app on first run, then click Open ScanContext.
Option B — winget

Install via winget

For developers and power users who prefer the command line. No SmartScreen warning — installs through Microsoft’s trusted channel.

winget install NikolaReljin.ScanContext
  1. Open PowerShell or Windows Terminal and run the command above.
  2. Open ScanContext from the Start menu when the install finishes.
Option C — no install

The download bundle (.zip)

No installer — just unzip and double-click a start file.

⬇ Download the bundle (.zip)
  1. Unzip it anywhere (Desktop, Documents…).
  2. Open the unzipped folder and double-click Start ScanContext.bat.
  3. First start downloads the app (a few minutes); then your browser opens ScanContext.
To stop it: click Stop in the app, or double-click Stop ScanContext.bat in the bundle folder.
Option A — recommended (Apple Silicon)

The ScanContext app

For Apple Silicon Macs (M1/M2/M3 and newer). On an Intel Mac, please use Option B below.

⬇ Download for macOS — Apple Silicon (.dmg)
  1. Open the .dmg and drag ScanContext into Applications.
  2. First launch: right-click the app → Open → Open (the app isn't code-signed yet — only needed once).
  3. It checks Docker, downloads the app on first run, then click Open ScanContext.
Option B — no install (any Mac)

The download bundle (.zip)

Works on both Apple Silicon and Intel Macs.

⬇ Download the bundle (.zip)
  1. Unzip it anywhere.
  2. Open the folder and double-click Start ScanContext.command.
  3. First start downloads the app (a few minutes); then your browser opens ScanContext.
To stop it: click Stop in the app, or run sh ./stop.sh in the bundle folder.
Option A — recommended

The ScanContext app

Pick the package for your distribution — or the portable AppImage.

⬇ AppImage (portable) ⬇ .deb (Debian / Ubuntu) ⬇ .rpm (Fedora / RHEL)
  1. AppImage: make it executable (chmod +x) then double-click it. .deb / .rpm: install it like any package.
  2. Open ScanContext. It checks Docker, downloads the app on first run, then click Open ScanContext.
Option B — terminal (always gets the latest)

One-line install with curl

Prefer the terminal? This always fetches the most recent release — no version numbers to track. Paste it into a terminal:

curl -fsSL "$(curl -fsSL https://api.github.com/repos/nikolareljin/scancontext/releases/latest \
  | grep -oE '"browser_download_url":[^,]*\.zip"' | cut -d'"' -f4)" -o ScanContext.zip \
  && unzip -q ScanContext.zip && cd ScanContext-* && sh ./start.sh

To download the AppImage instead, swap \.zip" for \.AppImage" in the command above.

Or download the bundle by hand:

⬇ Download the bundle (.zip)

First start downloads the app (a few minutes); then your browser opens ScanContext.

To stop it: click Stop in the app, or run sh ./stop.sh in the bundle folder.
Looking for a previous version, or a different file? Every release — current and older — with all of its downloads is listed here: all ScanContext releases ↗
Step 3 — you're in

Use ScanContext

When it's ready, ScanContext opens in your web browser. A short guided tour runs the first time — reopen help any time with the "?" button in the top-right corner. Import a study and explore. Your studies and data are always kept between runs.

📦 Browse all releases (incl. previous) ↗ Need a hand? See Help

The download buttons above always point to the newest release. Older versions are on the all-releases page.

Before you start

System requirements

You need

  • Windows 10/11, macOS 12 or newer, or a modern Linux desktop
  • Docker Desktop installed and running
  • About 4 GB free disk space, plus room for your studies
  • A modern browser — Chrome, Edge, Firefox, or Safari

Good to know

  • Internet is needed only for the one-time first download (and for cloud AI, if you enable it)
  • The app uses ports 5173 and 8000 on your machine
  • No account, no sign-up, no telemetry

Optional

Connect an AI provider

ScanContext works fully without AI. When you turn it on, it can summarize reports, translate non-English reports, and suggest regions of an image worth a closer look — always labelled "candidate findings, not a diagnosis". AI is off by default.

ProviderWhat it isPrivacy notes
Google GeminiRecommended. Free tier to try; paid tier for real use.Cloud. Anonymized data sent; paid tier required for genuine privacy.
Anthropic ClaudePaid cloud API.Cloud. Anonymized data sent to Anthropic.
OpenAIPaid cloud API.Cloud. Anonymized data sent to OpenAI.
OllamaFully local AI — no key, no internet.Nothing leaves your computer. Best for privacy.
Privacy & the Gemini free tier. ScanContext strips patient identifiers before sending anything to a cloud provider — but on Google Gemini's free tier, Google may still use your anonymized images and reports to improve its models. Only the paid tier stops that. For real medical data, use the paid tier — or use Ollama to keep everything on your machine.

Setting up Google Gemini

  1. Get a key at aistudio.google.com/apikey — sign in, click Create API key, copy it (starts with AIza…). Keep it private.
  2. Turn on billing (paid tier) in the key's Google Cloud project — see the warning above. You can set a budget cap.
  3. Add the key in ScanContext — no files to edit. Open ScanContext, click Settings in the top bar, then set:
    • AI provider: Gemini
    • Paste your key into Gemini API key (stored locally; the page never shows it back in full — only the last 4 characters)
    • Turn on Cloud AI enabled
    • Leave Anonymize before AI on (recommended)
    • Click Save settings
  4. Changes apply immediately — no restart needed. Then open a study and use Analyze study; you confirm cloud use per study.
  5. Your API keys are safe. Settings are protected automatically — only someone with access to your installation can change them. Keys are stored locally and never leave your machine.
Advanced — set the key in .env instead

For unattended / scripted setups, the same values can be written to the .env file. After editing it you must Stop, then Start ScanContext so it reloads.

AI_PROVIDER=gemini
CLOUD_AI_ENABLED=true
GEMINI_API_KEY=AIza...your-key...
GEMINI_MODEL=gemini-2.5-flash

Where the .env file lives:

How you installed.env location
Download bundle (.zip)Inside the unzipped ScanContext-<version> folder
App — Windows%APPDATA%\com.nikolareljin.scancontext\.env
App — macOS~/Library/Application Support/com.nikolareljin.scancontext/.env
App — Linux~/.local/share/com.nikolareljin.scancontext/.env

If both are set, the in-app Settings value wins.

Which Gemini model?

gemini-2.5-flash — cheap, fast, good for everyday use (the default).
gemini-2.5-pro — the most thorough analysis.

What does it cost?

Cost per study is capped by design (about a dozen sampled images). Roughly 2–3¢ per scan on Flash, 10–20¢ on Pro. Even 100 studies a month is about $2–3 on Flash. Rates are set by Google.

Help

Troubleshooting & common questions

It says "Docker is not installed" or "not running"

Install Docker Desktop, open it, and wait until it reports "running". Then start ScanContext again. On macOS, make sure Docker Desktop has been launched at least once.

Windows: Docker Desktop won't start (virtualisation not detected)

If Docker Desktop reports "failed to start because virtualisation support wasn't detected", it is a Windows / PC setup issue. Two independent fixes — most people need only one.

Fix A — turn on virtualization in your PC's BIOS/UEFI. Restart and press the setup key shown on the boot screen (often Del, F2, F10, or Esc). Find the virtualization setting — Intel: Intel Virtualization Technology / VT-x; AMD: SVM Mode / AMD-V — set it Enabled, save & exit, and reopen Docker Desktop.

Fix B — repair the Windows features and WSL. Open Turn Windows features on or off, untick Hyper-V and Virtual Machine Platform, click OK and restart. Then open Windows PowerShell and run wsl --update (if WSL is not installed, run wsl --install as administrator) and restart. Open Docker Desktop and click Try Again.

Docker's own guide: docker.com troubleshooting ↗.

The first start is taking a long time

The very first run downloads the app images — this can take a few minutes depending on your connection. Every start after that is fast.

My system warns the app is unsafe / unknown

The installers aren't code-signed yet, so you may see a warning the first time. macOS: right-click the app → Open → Open. Windows: on the SmartScreen prompt, click More info → Run anyway. You only do this once.

A port is already in use

ScanContext uses ports 5173 and 8000. If 5173 is taken, change FRONTEND_PORT in the .env file and start again. 8000 is built into the app and must stay free — close whatever is using it rather than changing BACKEND_PORT.

Is my medical data uploaded anywhere?

No — not by default. Everything stays on your computer. Data is only ever sent out if you turn on a cloud AI provider in Settings, and even then you confirm it per study. Use Ollama to keep AI fully local too.

Can ScanContext tell me what's wrong with me?

No. ScanContext is not a medical device and does not diagnose anything. It helps you organize and understand your files and prepare questions. Every result — especially anything AI-suggested — must be confirmed by a qualified clinician.

Where are my files stored?

In a data/ folder on your computer. With the bundle it's inside the unzipped folder; with the app it's in the per-user app data folder. Your studies are kept between runs — keep that folder and any exports private.

How do I stop it?

In the app, click Stop (window or tray / menu-bar icon). With the bundle, double-click Stop ScanContext.bat on Windows, or run sh ./stop.sh on macOS / Linux.

Still stuck?

Open an issue and describe what happened: github.com/nikolareljin/scancontext/issues.

Privacy & safety

What you should know

Safety

ScanContext is not a diagnostic tool and not a substitute for professional medical advice. Anything it shows — especially AI-suggested findings — is a prompt for a conversation with your doctor, never a conclusion.

Privacy

Medical files contain personal health information. ScanContext keeps them on your machine; cloud AI is opt-in and strips identifiers first. That anonymization only gives real privacy on a paid provider plan — on a free tier, use Ollama for genuine medical data.

Support the project

ScanContext is free — and built by one person

If it saved you time or worry, a one-time $10 donation keeps it maintained and improving. Any amount helps.

☕ Donate on Ko-fi