What is a digital twin, and what is a virtual twin?

What does a digital twin actually copy?

A digital twin is not a picture of something. It is a working model of something real, continuously fed with data from the original.

Digital twin technology was born in aerospace and manufacturing. Weather services run digital models of the atmosphere and update them every few minutes with readings from satellites and ground stations. Aircraft manufacturers run a digital twin of each individual jet engine, kept in sync with telemetry from every flight, so they can see how that specific engine is ageing rather than how engines age on average. The model earns its usefulness by never stopping: it is only ever as good as the last data it received.

A virtual twin applies the same principle to a person. What it copies is not your appearance and not your medical file. It copies the behaviour of your biology: the levels, the rhythms, the drifts, and the way they respond when something in your life changes.

What does the published research say about digital twins in health?

That the field is real, active and still early. The reference survey is Katsoulakis and colleagues, "Digital twins for health: a scoping review", published in npj Digital Medicine in 2024 (DOI 10.1038/s41746-024-01073-0, consulted 29 August 2026). It maps how the idea is being applied across medicine and finds a fast-growing but uneven literature, with most applications still at an early stage of evidence.

Two honest limits on that citation. A scoping review maps a field, it does not validate any single product, and it says nothing about BioTwin in particular. And a finding about a field is not a finding about you. What the review does establish is that the object described on this page is a recognised research subject, not a marketing invention.

Digital twin or virtual twin: which term is correct?

Both, and they are not interchangeable. BioTwin uses them for two different things, and keeping them apart is what makes each one precise.

Digital twin is the technology in general: a live model of a real system, whatever that system is. An engine, a production line, a power grid, the atmosphere. The term has an international definition, ISO/IEC 30173:2023, « Digital twin, concepts and terminology », and it is used across aerospace, manufacturing and energy.

Virtual twin is that same technology applied to a human being: a virtual copy of one person, built from that person's own biology.

In one sentence, which is the sentence worth remembering: digital twin technology is already transforming industry, and when BioTwin applies the same principle to a person, it calls the result a virtual twin. Same principle, different object, different purpose.

That distinction is BioTwin's own, and it is worth saying so plainly. The research literature uses several of these phrases interchangeably, and where it names the human application it more often says human digital twin or medical digital twin. BioTwin separates them because the separation is useful: it keeps clear what is a machine and what is a person.

So if you arrived here searching for a digital twin of your health, a personal digital twin, a digital health twin or a human digital twin, you are in the right place. Those phrases all point at the same object, and BioTwin treats them as synonyms of the term it leads with: virtual twin. Nothing about the method changes with the phrase you happened to type. What changes, and what this page is built on, is the distinction between the technology and its application to a person.

A virtual twin is not a scan sets out the distinction that matters even more than the vocabulary.

How is a virtual twin different from a health app or a results dashboard?

A dashboard displays what was measured. A twin models what was measured, which is a different kind of object and answers a different kind of question.

Three differences do most of the work.

Memory. A results sheet describes one moment. A twin keeps every previous moment and reads the new one against them. Your own history becomes the reference, not a population average.

Breadth. A panel reports a fixed list of selected analytes. A virtual twin is built to hold far more signals than any one report shows you, and to combine them into patterns rather than list them.

Direction of the question. A report answers what was my value. A twin lets you ask where is this going, and what moves when I change something. That is the question people actually have, and it is the one a static document cannot answer.

What data does a virtual twin need?

Five layers, and the fifth is the one people forget.

• Metabolomic: the small molecules circulating in your blood, which reflect what is happening in your body close to real time.
• Genomic context: what your biology tends to do.
• Phenotypic: measurable characteristics, including body composition.
• Behavioural: sleep, activity, nutrition, substances, the things you actually do.
• Environmental and temporal: where you live, how you travel, and above all how long the model has been watching.

The fifth layer is time. A twin built from a single sample is a good first photograph. A twin built from several samples across a year is a model. This is why your twin keeps learning matters more than any single result.

How does BioTwin build one?

It starts with a Bio-Signature sample collected at home, from a few drops of blood dried on a card. You post the kit back.

The sample is processed by high-resolution mass spectrometry, which separates the molecules present in the sample and measures them. That is how one small sample yields more than 30,000 biomarker-derived metabolic signals. Mass spectrometry, without the jargon explains the instrument itself, and the finger prick and what happens next follows the sample from your kitchen table to the result.

Those raw signals are then cleaned, normalized against BioTwin reference data so that they can be compared meaningfully, and aggregated. What comes out of that step is not a list of 30,000 numbers, which would be unreadable and useless. It is a set of Bio-Signatures.

What is a Bio-Signature, and why is that the output rather than the signal count?

Bio-Signatures are patterns derived from biomarker data that summarize a specific aspect of your biology, for example metabolic, inflammatory or nutritional signals, into an indicator you can actually follow over time.

The distinction matters. The number of signals is a means, not a promise. Thirty thousand signals that nobody can read are worth nothing. A Bio-Signature is what turns that volume into something a person can act on, and something a model can track from one sample to the next. Across BioTwin's medical and non-medical work there are practically hundreds of Bio-Signatures, and which ones you see depends on the program you are in.

In your BioTwin portal, Bio-Signatures are expressed as personalized scores covering the main axes of your biology. Each score is backed by hundreds of signals rather than a single line on a report. Complete 360° covers the lifestyle medicine pillars of Elite Health Optimization: fitness, nutrition, sleep, stress, risky substances, and social connection.

What can you ask a virtual twin that you cannot ask a report?

The next question. That is the whole difference, and it is worth being concrete about it.

A standard panel tells you about selected analytes at one moment. You can read it, and then you are done: the document has said everything it will ever say. A model of your biology lets you ask what happens if you change something. What if I sleep seven hours instead of five. What if I hold that training block for three months. What if I stop the thing I keep meaning to stop.

A report cannot take those questions, because it has no memory and no direction. A twin can, because the next sample is read against the ones before it. You are not comparing yourself to a population range that was never about you: you are comparing yourself to yourself, three months earlier, before you changed anything.

Why does a virtual twin need time before it becomes useful?

Because the first sample sets the baseline, and a baseline is not yet a direction.

One measurement tells you where you are. Two tell you which way you are moving. Several across a year tell you how fast, and whether what you changed had any effect on your biology rather than only on how you felt. Wearables measure signals continuously but from the outside. A Bio-Signature reads chemistry from the inside, less often, in more depth.

There is a second reason, less obvious. BioTwin retains the signals from your sample, so an analysis developed later can be run on a sample you already sent. Your twin can therefore gain depth without you sending anything new.

What can a virtual twin not do?

This part is short and it is not negotiable.

The BioTwin technology shown here is not authorized as a medical device. BioTwin is pursuing its regulatory plans with the FDA and Health Canada. Wellness experiences do not replace medical advice. TwinMe wellness outputs are non-medical wellness insights and are not medical or laboratory tests. They do not provide clinical concentrations, units, reference ranges, diagnoses, screening, treatment or disease monitoring. BioTwin's regulated clinical applications are separate from TwinMe and are available only in jurisdictions and clinical settings where authorized.

In plain language: a virtual twin will not tell you that you have a disease, and it is not built to. It will not replace your physician, and it will not hand you a lab report. What it does is describe how your own biology is moving, in language you can use, so that the conversations you do have with a clinician start from something better than a guess.

Is a virtual twin the same thing as a whole-body scan?

No, and the difference is a difference of object, not of quality.

An imaging examination looks at structure at one moment: shapes, densities, anything visibly present on the day of the examination. A virtual twin looks at function over time: chemistry, rhythms, and drift. Both are legitimate. They answer different questions, and one does not substitute for the other.

BioTwin does not position a twin as an alternative to imaging and does not recommend one over the other.

In June 2026 the Collège des médecins du Québec published its position on whole-body imaging in people without symptoms. It is a public document and worth reading. BioTwin does not put itself forward as the recommended alternative, because it is not one. Our conviction is simpler and it looks forward: artificial intelligence is going to help build a better twin, a great deal remains to be demonstrated on every side, and we would rather look at what each object is worth than settle the question in advance, including for ourselves.

A virtual twin is not a scan sets out the comparison in full.

Why a twin rather than a single test?

Because they answer different questions, and neither answer replaces the other.

If you want to know one specific value on one specific day, a single test is the right tool and it is very good at that. If you want to follow a state of health or wellbeing across months, a twin is the object built for that job. It is the difference between a photograph of a tree and a photograph of the forest: neither one is wrong, and they are not answers to the same question.

Who is a virtual twin actually for?

For someone who wants to see change rather than confirm a state.

In practice, the people who get most out of it are those who intend to keep going: who plan to test more than once, who already track something, and who want to know whether what they changed actually moved their biology.

How do you get a virtual twin?

Through TwinMe, the consumer platform where you build your twin and order at-home Bio-Signature kits.

Complete 360°, which covers the pillars of Elite Health Optimization, is $995 USD, or CAD for a Canadian shipping address. Standalone TwinMe access is CA$9.95 per month, and 12 months of access are included with any program. TwinMe is available across North America, in the United States and in Canada. Shipping is free in Canada, and international shipping is available. Results appear in your portal within 1 to 3 months of your sample arriving, and if you have sent more than one kit you also see how each score has moved since the previous one.

Order a kit on TwinMe