Med «notat»-funksjonen legger du til et ekstra lag med AI-drevet gjennomgang av dine genererte notater. Det analyserer notatet ditt mot kildedokumentene dine – transkripsjonen og klinisk kontekst – og flaggerer potensielle avvik, slik at du kan fange opp tingene som kan slipse igjennom under en manuell gjennomgang, spesielt i lange eller komplekse konsultasjoner. Du kan spore ethvert faktum tilbake til transkripsjonen, og løse problemer med ett klikk.
Det lærer du
How analyserer dine notater
How kjøre en verifisering og lese resultatene
How gjennomgå problemer og verifiserte enheter
How innebygde høydepunkter fungerer i notatet ditt
How bruke foreslåtte korrigeringer
How gi tilbakemelding om resultatene
Common questions and troubleshooting
Getting started med
After notatet ditt er blitt generert, vises en »notat« button i det nedre høyre verktøyraden i notateditoren din. Ved å klikke på det starter prosessen ved å kjøre en analyse som sjekker notatet mot kildedokumentasjonen din (transkripsjonen og klinisk kontekst) for å identifisere potensielle avvik eller utelatelser.
The analysis typically takes less than 2 minutes. Du trenger ikke å vente – du kan lukke panelet, navigere til andre økter eller fortsette å arbeide med andre dokumenter. You'll see progress updates as »notat« scans the note, checks entities, and identifies potential discrepancies. When the analysis is complete, you'll get a notification to view your results.
Running a »notat«
Start en verifisering
Click the »notat« button in the bottom-right toolbar of your note. The button changes to »notat«... and a panel opens showing progress. You can close the panel during analysis - it continues in the background. A notification will let you know when results are ready.
Re-run after editing
If you've edited your note and want to check it again, click the refresh icon in the panel header to re-run the analysis on the updated note.
»notat« resultater
When the analysis is complete, the verification panel opens with two tabs: »notat« and All »notat«.
»notat« works by extracting clinical »notat« from your note - these are the individual facts and details that make up a clinical record. Things like medication names, dosages, diagnoses, symptoms, procedures, vital signs, and patient demographics. Each entity is checked against the transcript and context to see whether it is accurately captured.
»notat« fane
This tab shows anything that may need your attention, grouped by priority:
High priority - could directly affect patient safety (e.g. potential issues with diagnoses, examination findings, lab results, or management plans; or critical clinical facts missing from the note)
Medium priority - clinically relevant but less likely to cause harm (e.g. potential issues in a referral, assumed patient age or gender, or contextual information missing from the note)
Low priority - not consequential to the clinical picture (e.g. minor details, small talk, or background information that doesn't affect clinical decision-making)
Each issue shows the entity name, its category, and what type of issue was detected. Click any issue to see more detail, including where it appears in the transcript.
If no issues are found, you'll see a confirmation that all content appears to be supported by the transcript with no missing entities.
All »notat« fane
This tab lists the clinical entities found in the note with their verification status. You can filter entities by category using the dropdown at the top of the tab.
Entity categories
Each entity is automatically classified based on its clinical type:
kategori | Examples |
«notat» | Drug names, dosages, frequencies |
diagnoser | «notat» vurderinger |
«notat» | Patient-reported complaints |
«notat» | «notat», «notat», undersøkelser |
Lab | »notat«, resultatene fra patologi |
vitale tegn | blood pressure, heart rate, temperature |
anamnese | Past medical history, family history |
demografi | Age, gender, occupation |
annet | «notat» som ikke passer inn ovenfor |
»notat«
Each entity links back to the part of the transcript or context it was identified from. Click View in transcript or View in context on any entity to see the original source text. This lets you quickly cross-reference the note against what was actually said in the consultation.
💡merk: »notat« checks that the note matches what is in the transcript. It does not verify the correctness of the transcript itself. If a medication name was misheard during the consultation (e.g. due to background noise), that error originates in the transcript and »notat« cannot catch it.
»notat« høydepunkter
Heidi Verify highlights entities directly in your note editor with a coloured marker, so you can see results in context without switching to the panel.
«notat» med problemer
Entities with potential issues are shown with an orange wavy underline. These are on by default so you can spot problems at a glance.
verifiserte «notat»
Verified entities are shown with a green highlight. These are off by default to keep your note clean - you can toggle them on from the All Entities tab if you want a full visual overview of what has been checked.
»notat« kontroller
Use the Highlight issues and Highlight verified switches in the panel footer to turn highlights on or off. Your preferences are saved for future sessions.
«notat« with highlights
Click a highlight in the note to select it in the panel and see full details
Click an entity in the panel to scroll to and highlight it in the note
Hover over a highlight to see a quick summary, including its source in the transcript
forslag «notat»
For some issues - such as potential discrepancies or content not supported by the transcript - Heidi Verify may suggest a fix.
Click Apply fix to update the note with the suggested replacement, then review before accepting. The entity briefly flashes green to confirm the change, and the issue is resolved.
For content that isn't supported by the transcript and should be removed entirely, click Remove from note to delete it.
You can undo any applied fix if needed.
Alternatively, if you prefer a different correction to those generated, you can always edit the note manually instead.
gi tilbakemelding
You can help improve Heidi Verify by providing feedback on individual results.
Thumbs up - confirms Heidi got it right (e.g. correctly verified an entity, or correctly flagged an issue)
Thumbs down - indicates Heidi got it wrong (e.g. incorrectly verified something, or incorrectly flagged something as an issue)
You can optionally add a text comment with more detail. Your feedback helps Heidi learn and improve over time.
FAQs
Where can I find the Verify button?
The Verify button appears in the bottom-right toolbar of your note editor after a note has been generated. If you don't see it, the feature may not yet be enabled for your account. Contact our Support team to find out more information.
Do I still need to review my notes manually?
Yes. Heidi Verify is not a replacement for your own review. The analysis may not catch every issue.
What types of issues does it flag?
Heidi Verify can flag several types of discrepancies, including: content in the note that doesn't appear in the transcript, information that contradicts the source, details from the consultation that are missing from the note, and supported inferences that are reasonable but weren't explicitly stated.
What does "supported inference" mean?
A supported inference is something that appears in the note but wasn't explicitly stated in the transcript - however, it's a reasonable clinical conclusion based on what was discussed. These are flagged for your awareness but aren't necessarily errors.
Can it catch errors in my transcript?
nei. Heidi Verify checks your note against the transcript. If the transcript itself contains errors (e.g. a medication name was misheard), Heidi Verify cannot detect that - it only flags when the note diverges from the source.
Can I verify a note I've edited?
Yes. If you've made changes to your note, click the refresh icon in the panel header to re-run the analysis. The new analysis will reflect your edits.
How long does the analysis take?
Typically under two minutes. You can navigate away, work on other sessions, or continue documenting while it runs. You'll get a notification when the results are ready.
What if a flag is wrong?
If you disagree with a flag, simply give it a thumbs down. Taking this action tells us the flag was incorrect, which helps improve accuracy over time. You can optionally add a comment explaining why.
Can I turn off the inline highlights?
Yes. Use the toggle switches in the footer of the verification panel to turn highlights on or off. Your preferences are saved for future sessions.
What if I don't want to use Heidi Verify?
Simply don't click the Verify button. The feature is entirely opt-in and never runs automatically. Your note workflow is unaffected if you choose not to use it.
Can I run Heidi Verify on any note?
Heidi Verify works best with notes generated from a live transcript. The richer the source data (transcript and clinical context), the more thorough the analysis.
Which plans include Heidi Verify?
Heidi Verify is currently in beta and available to some users on Practice and Enterprise plans. We're actively improving it based on your feedback. Please contact our Support team to find out more.