Explain My Report

It translates the medical language in a report into plain English. You get a short summary, a list of terms with a plain-language meaning and a supporting excerpt, plus questions you can ask your care team. It explains what the report says; it does not interpret what your future holds or tell you what treatment to choose.

No. It is an educational explanation of the document you provide. It is not a diagnosis, a second opinion, a treatment recommendation, or a substitute for a clinician who knows your history. If a finding worries you, contact your care team. For an emergency, use emergency services rather than this tool.

The text or image is sent to Ontool’s server and then to Google Gemini to produce the explanation. Ontool does not save the report or the result to a database, and there is no account or history. Like any online service, the request still travels through Google’s API and may be processed under Google’s terms. Remove names, addresses, and other identifying details before you submit if you do not need them.

Yes. PNG, JPG, and WebP images are accepted. Photos larger than 3 MB or wider than 2,200 pixels are resized in your browser before upload. A clear, straight, well-lit photo works best. Handwriting, glare, and cropped pages can make the model mark findings as unclear, and there is no source-text check for images because the readable text only exists inside the model’s reading of the image.

The quote is the evidence for the explanation. For pasted text, Ontool checks that each quote really appears in what you submitted and drops findings that fail the check. This is meant to stop the model from inventing context that is not in the report. The quote is an excerpt, not a complete medical record, and you should read it alongside the original.

They are simple labels, not a diagnosis or a severity score. Routine means the report presents the finding as expected, benign, or negative. Needs attention means the report flags it as abnormal or significant and your doctor should discuss it. Unclear means the tool could not confidently read or interpret that part of the report. All three labels are explanations of the report’s own language.

Without an account you can explain one report per day from the same connection (IP address). Signing in to a free account raises that to three explanations per day. The limit is tied to the connection or the account rather than a browser session, so refreshing the page, clearing cookies, or opening a new tab does not reset it. Failed model calls are returned to your allowance. The daily count resets at midnight UTC. The remaining count appears in the tool after each request.

Pasted text can be up to 20,000 characters. Images can be up to 3 MB after the browser prepares them, in PNG, JPG, or WebP. Very long reports should be split into the relevant section. The tool handles one report per explanation.