Privacy & Support
what is collected, and what never will beVery little is collected here, and two things never will be: nothing is sold, and nothing feeds advertising. There is no ad network and no tracking of you across the web. Every scan, atlas and tractogram is a file that ships with the page and is rendered on your own machine. The two optional services that do touch a network — an account, if you make one, and plain usage counts, if they are switched on — are named below, and the date at the bottom changes whenever this list does.
The apps
The iPhone, iPad, Mac and Android apps carry the whole site inside them — the MNI152 template, the AAL-116 and Brodmann atlases, the HCP1065 tractogram and the viewer. The anatomy makes no network requests. You can put the device in aeroplane mode and every page still works, which is a test you can run yourself. The network is used only by the optional services described below, and only when you use them.
The app stores no personal data. Two things are kept on the device and never leave it:
- Whether you chose the light or the dark theme.
- A snapshot, only if you ask for one. It is written into the app’s own documents folder and then handed to the system share sheet, so where it ends up is your choice. Nothing is sent anywhere.
On a Mac with Apple silicon the app is the iPad app, unchanged, running inside the sandbox macOS gives every app from its store. It asks for nothing beyond that sandbox: the only file it writes is a snapshot you ask for, into its own container, and the only connections it makes are the optional services below, when you use them.
The website
The website is the same files, served from Cloudflare. One thing follows from being on the open web rather than in an app bundle, and it is worth stating plainly:
- Cloudflare sees requests. Serving a page means the host receives your IP address and the usual request headers. This is how the web works and applies to every site you visit; nothing is added to it here, and none of it is used to build any profile.
The pages themselves make no third-party requests today. The typefaces used to be the exception — they were fetched from Google Fonts, which meant Google saw an IP address on every page load — and as of 27 August 2026 they are served from this site alongside everything else.
Your browser also stores, locally: the light-or-dark choice, and a cache of the site’s files created by its service worker so that it works offline. Both live in your browser, are readable only by this site, and are cleared whenever you clear site data.
Optional services
Two services can add to the list above, and they are worth naming before they exist rather than after:
- Questions asked through an AI assistant. The atlas answers questions inside Claude, ChatGPT, Kimi and other assistants that speak the Model Context Protocol, through a small server at mcp.visualneuroscience.ai. That server answers from the same public data as these pages, holds no accounts, stores nothing about you, and keeps no record of your questions beyond the short-lived operational logs of the platform it runs on. What you type to the assistant itself is governed by that assistant’s own privacy policy, not this one.
- An account, if you make one. When a Log in button is present, sign-in is handled by Google’s Firebase service. It stores your email address, your name, the username you choose (which is shown in place of your name, and which no one else can take), your university if you give it, the files you choose to save (such as the named models of the Region Atlas), and your progress in the practice questions, so that the same account works on the web and in the apps. That data exists to serve your account back to you: it is deleted with the account, and it is never sold or used for advertising.
- Usage counts. The site may count how its pages are used — which tools are opened, on what kind of device — to decide what to build next. If and when such counting runs, this page will name the provider and what it records. It would measure use of this site, not you across the web.
Deleting your account
Open Log in on any page and, signed in, press Delete account: the account, the files saved to it and its practice progress are removed at once, and nothing of them is kept. If you would rather not do it from the app, write to fkarim@visualneuroscience.ai from the address on the account and it is deleted by hand, with the same result. Deleting the account does not remove you from the newsletter, which is a separate list: every newsletter email carries its own unsubscribe link.
Children
The site is an educational anatomy reference. It is not directed at children and does not knowingly collect personal information from them.
What this is not
Nothing here is a medical device, a diagnosis, or advice about treatment. The pages summarise published work and draw published atlases; each says where its data came from and where the evidence stops. Read them as a reference, not as a clinical tool.
Changes
If this policy changes, the date below changes with it. The Google Fonts disclosure that used to sit above is the kind of change to expect: something removed, with the date to show when.
Support & contact
Questions, corrections and bug reports are welcome — particularly corrections. The pages make factual claims about neuroanatomy and cite their sources, and if one of them is wrong I would rather know.
Write to fkarim@visualneuroscience.ai and a person reads it. VisualNeuroscience.AI is made by OttomanLabs.
Last updated 28 September 2026 · VisualNeuroscience.AI · OttomanLabs