Introduction
This guide walks you through training the software to match your personal writing style for clinical notes and formal documents. By providing a sample of your past work, the system learns your preferred phrasing and tone. This helps ensure that future AI-generated reports sound just like you, saving you time on editing.
1. Select a writing style to train
Choose whether you want to train Audzone for your clinical notes or formal documents. In this example, 'Document Writing Style' is selected.
/settings/profile
2. Paste a sample of your writing
Click into the text box and paste an example of a previous report or note that reflects your preferred writing style. As soon as you paste your example note or document, analysis will automatically start. It takes about 30 seconds to complete.
Note: Please do not paste empty templates here; we need an actual example of a note or report written by you for this to work.
/settings/profile
3. Review extracted clinical phrases
Once the analysis is complete, the system may identify specific clinical phrases from your sample. As a bonus, it also captures the voice you write your notes and reports in. This will be used to make future AI notes and reports sound more like you.
You can expand phrase categories, such as History, to review how you typically explain results. Just remember, any clinical phrases will always use the language you define in these steps. Decide whether you prefer AI on-the-fly language (usually better for history) or your specific language use via phrases, which is good for sections like Otoscopy, Tympanometry, and Audiometry, where you usually describe results using specific language.
/settings/profile
4. Deselect unwanted phrases
Review the extracted phrases and untick any that you do not want to save to your profile. In this instance, we untick all History phrase examples as we'd like the AI to use history descriptions inferred directly from the transcript in a session. This means our notes don't all read the same for the history section, for instance.
/settings/profile
5. Review phrases in other categories
Scroll down to see phrases extracted for other sections, such as Otoscopy. Click on any category to expand it and review the suggested wording.
/settings/profile
6. Save your selected phrases
Once you are happy with the phrases the AI has extracted, click the 'Save' button. This adds your preferred language to your profile for future use.
/settings/profile
7. Add more variations to a category
If a category only has one or a few phrases, click 'Add more variations' to let the AI generate additional ways to describe different clinical situations based on your writing style.
/settings/profile
8. Generate additional phrases
Click "Complete Category" to have the AI generate more phrases based on your writing style. You can also provide optional instructions to guide the AI.
/settings/profile
9. Review and edit generated phrases
The AI will generate a list of new phrases based on your writing style. You can review these and edit the text of any phrase to better suit your needs.
/settings/profile
10. Save the generated variations
After reviewing and editing the newly generated phrases, save the selected variations to your profile. (Depending on the size of the list, it might be hidden slightly under the support button. The support button can be dragged to another position)
/settings/profile
11. Go to your saved phrases
Click 'My Phrases' in the left-hand menu to view and manage all the phrases you have saved.
/settings/profile
12. Review your generated phrases
Click on a category, such as Audiometry, to view the phrases saved within it. You can now see all the newly generated phrases listed under the category. You can also edit them, add more categories, or extract more examples from here.
Now whenever you generate an AI note or report, your language will be used from your phrase library, reducing your need for edits. Make sure to provide examples of both a note and a report for this to work for both notes and reports in a session.
/settings/profile