Transcription and dictation
Myukura uses voice in two different ways:
Transcription, to turn a clinical conversation into a structured, reviewable base.
Dictation, so the professional can add, complete or correct information directly where it is needed.
Everything remains editable, reviewable and under professional control.
Clinical voice in Spanish · speaker separation · editable dictation · human review
Two ways of working with voice
Conversation and dictationserve different purposes
Not all clinical information appears in the same way.
Some of it comes up during the conversation with the patient.
Some comes afterwards: an observation, a clarification, a detail the professional wants to add or a note they prefer to record by voice.
That is why Myukura separates both workflows from the start.
Designed for conversations between several people.
Captures what happens during the consultation and keeps the structure needed to work on that information afterwards.
Designed for the professional’s direct input.
Lets you add new information by voice without having to reconstruct a whole conversation.
Consultation transcription
From a natural conversationto a base that can be reviewed
During a consultation, symptoms, history, medication, progress, doubts, decisions and next steps come up.
Myukura turns that conversation into text so it can be used as the basis of the documentation workflow.
But a clinical transcription shouldn’t be limited to producing a list of words.
It needs to keep enough context to be able to understand afterwards:
The goal is to reduce the work of reconstructing a consultation once it is over.
Speaker diarisation
Knowing what was said also means knowing who said it
The same sentence changes meaning completely depending on who says it.
It is not the same:
“I’ve been taking ibuprofen since Monday.”
“You can take ibuprofen.”
The words may be similar.
The context is not.
That is why Myukura’s transcription includes diarisation, the process that separates the different turns of a conversation and associates them with different speakers.
This separation helps preserve the structure of the consultation before the information passes to the following layers of processing and review.
Diarising does not mean biometrically identifying a person. It means distinguishing who spoke and when within a conversation.
More than speech-to-text
The structure of the conversationmatters too
In a real consultation, you find:
The quality of a clinical transcription doesn’t depend solely on recognising words correctly.
It also depends on preserving the organisation of the conversation.
Myukura uses that structure as one of the bases for working on the clinical information afterwards.
Clinical voice in Spanish
Designed aroundhow people really speak
An important part of Myukura’s development focuses on speech recognition applied to the clinical context in Spanish.
We work on elements that are especially error-prone, such as:
During our participation in the 8th MareNostrum Hackathon at the Barcelona Supercomputing Center, we worked on clinical speech recognition in Spanish and the evaluation of specialised systems.
The goal is not for a technology to work only under ideal conditions.
It is to understand how it responds to the language that actually occurs in a consultation.
Clinical dictation
Add informationwithout going back to the keyboard
Not everything the professional needs to document is said during the conversation.
After the consultation, the need may arise to:
That is what dictation is for.
The professional can use their voice to enter new information directly into the Myukura workflow.
Dictate whereyou need context
Dictation can be used as an additional form of input in different parts of the documentation work. It can be used to:
Add information the professional considers necessary after reviewing the consultation.
Add a detail that didn’t come up during the original conversation.
Enter new follow-up information.
Speak instead of typing when a rigid structure isn’t suitable.
Add information that improves or completes the existing documentation.
Editable at all times
Voice doesn’t turn anythingautomatically into final
Dictation is a form of input.
Not a decision.
Once converted to text, the professional can:
Edit the content directly.
Adjust terms, figures or context.
Add new information.
Discard whatever should not remain.
Check the content before adding it to the documentation workflow.
Voice reduces friction.
Control remains human.
Context traceability
Knowing what information appearsand where it comes from
An original conversation, a later contribution from the professional and an AI-generated structure are not exactly the same thing.
Myukura is designed to keep the origin of information distinct within the workflow.
Content may come from:
Information captured during the consultation.
Information added directly through the professional’s voice.
Content written by the user.
Information organised or proposed by the platform’s systems.
This distinction helps the professional during review.
It doesn’t only matter what appears.
It also matters how it got there.
From voice to documentation
A continuous workflow,but supervised
Voice is the entry point.
Then the documentation work begins.
Transcription preserves what happened.
Dictation adds new context.
Myukura organises the information.
The professional decides what should continue.
The conversation happens once
Documentation can bereviewed at leisure
During a consultation there isn’t always time to arrange every detail.
Transcription helps preserve what happened.
Dictation lets you complete afterwards whatever the professional considers necessary.
This way the consultation can end without all the documentation having to be settled at that very moment.
Under professional control
Myukura uses voice as a capture and documentation tool.
Not as a substitute for clinical judgement.
A transcription may need corrections.
A dictation may need editing.
Structured information may need review.
That is why the professional stays within the process.
Summary
Transcription and dictationinside Myukura
Clinical conversation turned into a reviewable text base.
Separation of turns and speakers within the conversation.
Direct input from the professional to add information by voice.
Content can be edited and corrected.
The origin of the information remains distinct within the workflow.
The final information remains under professional control.
Voice to document, not to lose control
Voice technology can reduce a significant part of the work between a consultation and its documentation.
But it only adds value if it keeps clear what happened, what was added afterwards and what the professional has reviewed.
That is Myukura’s approach.
Myukura is currently still in development and testing. The capabilities described on this page may evolve as the product and the evaluation of its different components progress. Myukura is an administrative and documentation support tool and does not replace the clinical judgement, diagnosis or decision of a healthcare professional.