Guides · Chrome extension

RadDictate Chrome Extension

The extension puts dictation inside the reporting system you already use — any web-based RIS or HIS that opens in Chrome. You speak; the text lands where your cursor is.

Contents 1 · Install & sign in 2 · Dictating into your RIS 3 · The on-page pill & F8 4 · Dictate on phone 5 · Grab text & proofread 6 · History: copy & transfer 7 · Audio backups (REC) 8 · Templates (TPL) 9 · Troubleshooting

1 · Install & sign in

Install it from the Chrome Web Store — one click, and it keeps itself updated:

  1. Open the RadDictate listing on the Chrome Web Store and click Add to Chrome.
  2. Click the puzzle-piece icon in the toolbar and pin RadDictate.
Manual install (only if the Web Store is blocked on your workstation)
  1. Download the extension from your setup page and unzip it to a folder you'll keep (Chrome loads it from there — don't delete it).
  2. Open chrome://extensions, switch on Developer mode (top right).
  3. Click Load unpacked and pick the unzipped folder.
  4. Click the puzzle-piece icon in the toolbar and pin RadDictate.

Manual installs don't auto-update — prefer the Web Store where possible.

Click the icon and sign in with your email and password. Signing in also unlocks your end-to-end encrypted history — the same one-password rule as everywhere else (the mobile guide, §2, explains why the email reset is destructive).

SCREENSHOT · img/extension/01-popup-signin.png — popup sign-in
The popup before sign-in.

2 · Dictating into your RIS

The golden rule: text lands where your cursor is.

  1. In your RIS, click into the box you want the report in.
  2. Open the popup and press ⏺ START DICTATION (or just press F8 / Numpad + — see §3).
  3. Dictate. Finish. The structured report is inserted at your cursor — mid-sentence if that's where you clicked.

On reporting systems with separate fields (e.g. APEX-style forms), RadDictate goes one better: findings are inserted into the findings box and the conclusion into the conclusion box, automatically.

SCREENSHOT · img/extension/02-report-tab.png — REPORT tab
The REPORT tab: dictate, phone, upload, proofread.
SCREENSHOT · img/extension/04-apex-insert.png — findings/conclusion split (demo form)
Findings and conclusion landing in their own fields.
When the split does — and doesn't — happen The findings/conclusion split only fires when both boxes are empty at the moment you start dictating. If either box already has text in it (a template you loaded, notes you typed), the split is skipped and your whole dictation is inserted as one block into the box your cursor is in instead — nothing is overwritten. This rule is the same everywhere text gets inserted: live dictation, 📁 UPLOAD AUDIO, Dictate on phone, and History Transfer all route through the identical insertion logic.

Also on the REPORT tab: 📁 UPLOAD AUDIO transcribes an existing recording instead of live dictation.

3 · The on-page pill & F8

While dictating, a small floating pill appears on the page — your recording status (REC / PAUSED / NO SIGNAL) and controls, draggable anywhere. You don't need the popup open while you dictate.

SCREENSHOT · img/extension/03-dictation-live.png — pill while recording
The pill: dictation control without leaving the report.

4 · Dictate on phone

No microphone on the workstation — or one you wouldn't put near your mouth? Press 📱 DICTATE ON PHONE. Your phone (with the mobile app signed in) pops up a record prompt; you dictate on the phone, and the report arrives back in the browser and inserts exactly like live dictation — same cursor-target and findings/conclusion split rule as §2. The result travels end-to-end encrypted.

SCREENSHOT · img/extension/09-phone-dictate.png — waiting for phone
The PC waiting while you dictate on the phone.

5 · Grab text & proofread

🔍 GRAB TEXT & PROOFREAD reads the report already on the page and gives it a second read: laterality slips ("right" in findings, "left" in conclusion), anatomic-site mismatches, internal contradictions. Issues are flagged and a suggested corrected version is offered alongside — nothing on the page is changed for you; you review and copy only what you accept. Use it on dictated reports, typed reports, anyone's reports.

6 · History: copy & transfer

The HISTORY tab lists your end-to-end encrypted reports from every device — the study you dictated on the phone in the car is here waiting. Tap a report to open it right there:

History auto-deletes after 14 days by default; the retention setting (1–60 days) lives on this tab.

Clear is cloud-wide ✕ CLEAR HISTORY deletes your synced reports from the cloud — they vanish from your phone and desktop too, permanently.
SCREENSHOT · img/extension/05-history-list.png — history list
Cross-device history.
SCREENSHOT · img/extension/06-history-detail.png — report detail with COPY/TRANSFER
Open in place; copy or transfer.

7 · Audio backups (REC)

The 🎙 REC tab keeps local copies of your dictation audio — filter RAW / COMPACTED, export with 📦 ARCHIVE, remove with 🗑 DELETE.

Retranscribe Every recording card has a ↻ Retranscribe button — it re-runs that saved audio through the same pipeline as 📁 UPLOAD AUDIO and saves the result to the REPORT tab and History as "Retranscribed: <name>". Unlike mobile, the extension doesn't auto-retry a failed live dictation — instead the on-page pill offers RETRY/SAVE mid-flow (§3), and REC-tab Retranscribe is the manual redo for any saved recording afterward. Between the two, a dropped connection never costs you the audio.
Local-only Like on mobile, recordings are never uploaded. Clearing the browser profile or removing the extension deletes them permanently — ARCHIVE first if you want to keep any. Signing in as a different user on the same computer also clears these recordings, so they are not exposed to the next account; your saved reports are unaffected and re-sync from your encrypted cloud history.
SCREENSHOT · img/extension/07-rec-tab.png — REC tab
Local audio backups.

8 · Templates (TPL)

Same templates, everywhere: the 📋 TPL tab syncs with your account, so a template built here works on the phone and desktop too.

SCREENSHOT · img/extension/08-tpl-tab.png — template editor
Templates: name is the trigger.

Two worked examples

A straightforward one. Create a template named "Normal Chest" with the body of a standard normal chest X-ray report, and add "CXR normal" as an Advanced trigger. Say either "normal chest" or "CXR normal" at the start of a dictation, and the AI builds the report on that template instead of transcribing from scratch.

When something you say doesn't fit. Say you dictate onto a template but mention a finding the template has no place for. The AI won't guess where to put it; instead it pulls that piece out and shows it in a persistent toast with a COPY button, so nothing you said is lost — copy it and add it to the report by hand. This is the honest answer to "I dictated something extra and it didn't make it into the template" — deliberate safety behavior, not a bug.

9 · Troubleshooting


Something this guide doesn't answer? Contact us or use the feedback form.