Before the visit
Prepare a timeline from the materials the patient brings. Keep dates, questions and missing information visible.
NEUROLOGY · CLINICAL AI · CONTINUITY OF CARE
Understand the patient. Connect the evidence.
A research platform for neurological and multidisciplinary care. NeuroCifra brings documents, consultations and follow-up into one clinical context — with AI assistance and a specialist in control.
01 / Our approach
One person. Too many disconnected pieces.
A referral in a PDF. A symptom mentioned during a visit. A result from last year. The challenge is to understand how they fit together. Our work focuses on making that connection visible.
During the conversation
Explore speech-assisted notes and structured review so important details can be revisited and checked.
After the consultation
Connect agreed actions, observations and follow-up. Compare change against the previous visit.
02 / ILLUSTRATIVE WORKSPACE
See how a consultation comes together.
Choose a stage to explore the proposed workflow. The preview is illustrative and contains no patient information.
01 / Material
Before the visit
Prepare a timeline from the materials the patient brings. Keep dates, questions and missing information visible.
PDFMRI / CTNOTES
02 / Consult
During the conversation
Explore speech-assisted notes and structured review so important details can be revisited and checked.
VOICECONTEXT
03 / Specialist review
Can AI preserve clinical meaning?
Evaluate documents and speech for lost negations, ambiguous doses, dates and missing context. A readable summary must also be faithful.
SOURCEREVIEW
04 / Next step
After the consultation
Connect agreed actions, observations and follow-up. Compare change against the previous visit.
VISIT 01VISIT 02
↗ WHAT THIS MEANS FOR PEOPLE
For the specialist: a clearer starting point and traceable information. For the patient: an explanation they can return to and a next step they can understand.
03 / Specialties
Three research questions. One clinical purpose.
We study whether AI can preserve the meaning of clinical information, whether visual explanations improve understanding, and whether connected records support a more coherent consultation.
Can AI preserve clinical meaning?
Evaluate documents and speech for lost negations, ambiguous doses, dates and missing context. A readable summary must also be faithful.
Can a picture improve understanding?
Explore the relationship between anatomy, observations and explanatory models. Keep measured results distinct from illustrative hypotheses.
Can we make change understandable?
Connect repeat assessments and everyday goals. Distinguish a measured change from a causal claim about treatment.
04 / Inside our research
Inside our research
Research perspectives on clinical context, documents, speech, visual explanation, follow-up and child development.
NEUROLOGY · enClinical context: why a symptom is not a diagnosis
What must an AI assistant preserve when it connects complaints, examination findings and a patient’s history? We examine the difference between recognising a pattern and establishing a clinical conclusion.
A complaint becomes clinically meaningful through context. The same description of forgetfulness can refer to different difficulties: sustaining attention, learning new information, retrieving a name or organising everyday activity. A useful clinical record must preserve these distinctions rather than replacing them with a single broad label. This is the starting point for our work on connected clinical information.
Explore the article ↗
CLINICAL PRACTICE · enFrom medical documents to a reliable patient history
Recognition is only the first step. Dates, medication changes, uncertain readings and the provenance of every statement determine whether a summary can support a consultation.
A medical document is more than a block of text. Its meaning depends on who produced it, when it was written, what examination it describes and whether a recommendation was later changed. A discharge letter, a referral and a patient’s note can contain similar terms while serving different purposes. Document processing must preserve these relationships before it can produce a useful clinical summary.
Explore the article ↗
SPEECH & CONSULTATION · enSpeech during a consultation: preserve meaning, not just words
Near-real-time transcription raises a clinical question: how do we preserve the speaker, negation, correction and uncertainty while the conversation is still unfolding?
A consultation is not a dictated report. People interrupt themselves, change a number, return to an earlier event and speak over one another. A patient may quote another person, while a specialist may offer a hypothesis as a question. Converting this exchange into a fluent paragraph can erase distinctions that are essential to understanding the encounter.
Explore the article ↗
VISUAL UNDERSTANDING · enThe brain as an explanation: what a visual model can show
An anatomical atlas, a measured imaging finding and a functional hypothesis communicate different kinds of knowledge. A useful visual interface must make those differences visible.
Visualisation can organise a difficult conversation. It offers a shared object that a specialist and a patient can examine together: an anatomical structure, the location described in a report or the course of an observation. Its value depends on whether the image clarifies the evidence rather than making an uncertain interpretation appear more certain.
Explore the article ↗
CONTINUITY OF CARE · enFollow-up: distinguish a change from an explanation
Longitudinal care requires more than a rising or falling chart. We examine comparability, missing observations, patient goals and the limits of attributing change to treatment.
A follow-up consultation asks a different question from the first visit. It asks what changed, under which conditions and whether that change matters to the patient. A score, a description of daily functioning and an account of treatment experience may each contribute something different. A useful longitudinal record should keep them connected without flattening them into one number.
Explore the article ↗
CHILD DEVELOPMENT · enChild language development: from parental concern to an assessment plan
Connecting family observations, language assessment and professional support while keeping digital assistance separate from premature diagnosis.
“My child understands everything but says very little” starts a conversation; it does not settle it. Concrete examples help: asking for help, following an unfamiliar instruction, telling a story or joining a game. They make the concern more specific than a general developmental label.
Explore the article ↗
LEARNING & DEVELOPMENT · enAttention and learning: why one questionnaire is not enough
Discussing school difficulties without labels, comparing home and classroom observations, and defining a useful role for AI in consultation preparation.
A child may lose track of instructions, take time to begin work or avoid reading. These descriptions concern different tasks. A useful review starts by asking where the difficulty occurs, what the child can already do and which assistance changes task performance.
Explore the article ↗
PAEDIATRIC NEUROLOGY · enPaediatric neurology and developmental education: one shared pathway
Connecting medical questions, developing skills and family goals while preserving each professional’s responsibility.
A child may see several professionals while the family retells the same history at every appointment. A shared record should connect consultations: which question was raised, what has been assessed and what the next team member needs to know.
Explore the article ↗05 / Connect
Let’s start a conversation.
For specialists, research teams and organisations exploring thoughtful applications of AI in care. Tell us about your interest.
Research version. Clinical decisions remain with qualified professionals.