Rad AI Reporting offers three powerful AI features to enhance radiology report creation: Omni Report, Omni Impression, and Omni Unchanged. This article introduces each feature, how to access them, and when to use them depending on your workflow needs.
In this article:
AI features overview
Comparison table
| Feature | Scope | Ideal for |
| Omni Report | Full report generation (findings and impression) | New or complete reports |
| Omni Impression | Impression section only | Summarizing or revising impressions |
| Omni Unchanged | Merges prior and current findings, inserting into appropriate sections | Follow-up reports with new or unchanged findings |
Omni Report: Full report generation
Omni Report is designed for comprehensive report generation. It populates your entire report—including relevant sections and the impression—based on your dictated findings.
Key capabilities
- Populates all report sections from a single, free-form dictation
- Ensures structured formatting and standardized language
- Replaces picklists with relevant findings
- Removes contradictory negative language
- Uses blue text to visually distinguish AI-inserted content from dictated and typed text
When to use
- You're starting a new report from scratch.
- You want the AI to generate all sections at once.
Trigger
Within the report editor, perform one of the following actions:
- (Recommended) Say "Omni Report".
- Click the Omni Report button in the Omni Box.
- Press your assigned hardware button or keyboard hotkey.
Omni Impression: Impression-only generation
Omni Impression focuses solely on generating the impression section of your report. It converts your summarized findings and overall conclusions into a clean, structured impression.
Key capabilities
- Drafts or revises the impression only
- Is customized based on the report context and the learned style of each radiologist
- Can be used multiple times if the impression needs editing
- Uses blue text to visually distinguish AI-inserted content from dictated and typed text
When to use
- You're generating or revising an impression.
- You’ve already populated other sections in the report manually or via a template.
Trigger
Within the report editor, perform one of the following actions:
- (Recommended) Say "Omni Impression".
- Click the Omni Impression button below the Omni Box.
- Press your assigned hardware button or keyboard hotkey.
Learn more about Omni Impression.
Omni Unchanged: Efficient prior comparisons
Omni Unchanged allows for free-form dictation of just the changed findings between a prior comparison and the current study—saving time and effort. Findings from the prior report and newly-dictated changes will be inserted into the relevant sections of the current report.
Key capabilities
- Parses and merges findings from both the prior report and the current Omni Box content
- Automatically inserts findings into the most relevant section of the current report
- Retains the original language from the prior report while applying the current template structure
- Removes outdated phrasing (e.g., “since prior study,” “new”) that no longer applies
- Prioritizes current dictation if it conflicts with prior report content
- Populates comparison and history fields based on the selected prior report
- Uses blue text to visually distinguish AI-inserted content from dictated and typed text
When to use
- You're documenting a follow-up study with findings that are unchanged from a prior study.
- You want to save time by auto-inserting stable findings without re-dictating everything.
- You need to preserve the original phrasing from a previous report while updating it to match the current report structure.
- You’ve already dictated new findings and want AI to merge them with the unchanged findings from a prior report.
Trigger
Within the report editor, perform one of the following actions:
- (Recommended) Say "Omni Unchanged".
- Press your assigned hardware button or keyboard hotkey.
- Click the Omni Unchanged button in the Omni Box.
Learn more about Omni Unchanged.
Accessing Omni Features
Turning on the Omni view
Omni is turned on by default. If you don't see the Omni view in your report workspace, here's how to turn it on:
-
Click on your name at the bottom of the left-hand navigation menu. Then select Personal Settings.
-
Go to Appearance > Workbench > Enabled views.
-
Make sure the toggle for Omni is turned on.
Using the Omni Box
All three features are available via the Omni Box in the report workbench:
-
Open a report.
-
Apply a relevant template to the report.
-
Navigate to the Omni Box by performing one of the following actions:
-
With your microphone on, say the voice command: "Omni Box".
💡 To leave the Omni Box, use the voice command “exit Omni”. Learn more about voice commands. - Click into the Omni Box.
-
Press the stop/play button on your microphone (e.g., PowerMic/SpeechMike) or your assigned button if changed from the default.
💡 If you're already in the Omni Box, the same button will toggle focus back to the report editor. You can customize your hardware button assignments and hotkeys (keyboard shortcuts) in Personal Settings.
-
-
When the Omni Box is highlighted in green, it's in focus and ready to capture dictation.
-
Turn on your microphone, switch your attention to the images in your PACS, and dictate your findings in free-text form.
-
Once done, activate the relevant Omni feature by either:
- (Recommended) Using a voice command.
- Clicking the Impression, Report, or Unchanged button below the Omni Box.
- Pressing the assigned hardware button or keyboard hotkey.
Tips and best practices
-
Review AI content: Always review the generated text to ensure clinical accuracy.
-
Optimize your templates: Follow these template best practices to get the most out of Omni.
-
Combine features: Use Omni Report first. Then revise with Omni Impression.
-
Use substitutions: Add your own auto-replace rules to ensure consistent formatting, refine the speech engine's text output, or create content shortcuts. Learn more about substitutions.
- Report issues: Use the in-app feedback dialog to report any recurring accuracy or performance issues with our AI features.
Learn more
For feature-specific walkthroughs and best practices, view the following articles: