TL;DR: Key Takeaways
- Personalisation is happening at both ends: monitoring cadence and treatment design
- AI's proven role today is prioritisation and consistency, not diagnosis
- Personalised vaccines remain investigational, with a strong randomised signal in melanoma
- Stage at diagnosis still dominates outcomes — early detection keeps its value
- Clinical responsibility stays with clinicians at every stage
Two ends of the same journey
At the early end, personalisation means your risk factors deciding how often you check, which lesions you photograph and how quickly you escalate. At the treatment end, it means therapies constructed from an individual's own tumour sequencing. They rely on different technologies and carry very different levels of evidence, but both are described as "personalised", which is why the terms are so often blurred.
Stage by stage: now and next
| Stage of care | Where AI is used today | Plausible next step |
|---|---|---|
| Public awareness | Plain-language risk education that reaches people before a lesion changes. | Personal risk profiles that change how often someone is prompted to check. |
| Screening | Low-friction photo screening that prompts action sooner rather than replacing a clinic visit. | Better calibration across skin tones and lesion types, published per-subgroup. |
| Clinical triage | Helping services prioritise the people most likely to need urgent assessment. | Structured photo histories accepted directly into referral pathways. |
| Dermatologist assessment | Structured photo history handed to the clinician at the point of examination. | Side-by-side change measurement presented as part of the consultation record. |
| Lesion monitoring | Consistent, comparable images over months instead of memory and guesswork. | Change detection that flags drift a person would not notice unaided. |
| Genomic analysis | Tumour and germline sequencing interpreted at a scale manual review cannot match. | Faster turnaround from biopsy to a usable target list. |
| Treatment selection | Ranking candidate targets for investigational, individualised therapies. | Better prediction of which patients respond, not only which targets to include. |
| Recurrence monitoring | Follow-up imaging and tracking after treatment, under clinical supervision. | Combined imaging and blood-based signals reviewed on one timeline. |
Public awareness
- Where AI is used today
- Plain-language risk education that reaches people before a lesion changes.
- Plausible next step
- Personal risk profiles that change how often someone is prompted to check.
Screening
- Where AI is used today
- Low-friction photo screening that prompts action sooner rather than replacing a clinic visit.
- Plausible next step
- Better calibration across skin tones and lesion types, published per-subgroup.
Clinical triage
- Where AI is used today
- Helping services prioritise the people most likely to need urgent assessment.
- Plausible next step
- Structured photo histories accepted directly into referral pathways.
Dermatologist assessment
- Where AI is used today
- Structured photo history handed to the clinician at the point of examination.
- Plausible next step
- Side-by-side change measurement presented as part of the consultation record.
Lesion monitoring
- Where AI is used today
- Consistent, comparable images over months instead of memory and guesswork.
- Plausible next step
- Change detection that flags drift a person would not notice unaided.
Genomic analysis
- Where AI is used today
- Tumour and germline sequencing interpreted at a scale manual review cannot match.
- Plausible next step
- Faster turnaround from biopsy to a usable target list.
Treatment selection
- Where AI is used today
- Ranking candidate targets for investigational, individualised therapies.
- Plausible next step
- Better prediction of which patients respond, not only which targets to include.
Recurrence monitoring
- Where AI is used today
- Follow-up imaging and tracking after treatment, under clinical supervision.
- Plausible next step
- Combined imaging and blood-based signals reviewed on one timeline.
What has to be proven before any of this is routine
Open questions
- Whether screening tools perform consistently across skin tones, lesion types and image quality.
- Whether individualised therapies help broad patient populations, not selected trial cohorts.
- Whether per-patient manufacturing can run at the speed and cost health systems need.
- Whether the computational component adds measurable value independent of the rest of the chain.
What you can do now
- Learn your personal risk factors — family history, skin type, sun and burn history
- Protect against UV; it remains the largest modifiable factor
- Check your skin on a set cadence rather than only when worried
- Photograph anything you are watching so change is visible, not remembered
- Escalate new, changing, bleeding or non-healing lesions to a clinician
Start with your own risk picture
Work through the factors that decide how closely you should be watching your skin.
Where ScanSkinAI sits
We work at the awareness, screening and monitoring end only. Our measured performance and its limits are published in the accuracy report and methodology paper. We do not diagnose, treat or provide access to any investigational therapy described in this cluster.
Frequently Asked Questions
No. Image-based AI produces a risk-awareness output, not a diagnosis, and it holds no clinical responsibility. Diagnosis is made by a clinician, normally after dermoscopic examination and, where indicated, a biopsy with laboratory histopathology.
It describes care shaped around one individual: their personal risk factors and monitoring cadence at the early end, and — in research settings — therapies designed from their own tumour sequencing at the treatment end.
No timeline can be responsibly stated. Programmes are investigational, the most advanced has reported topline Phase 3 endpoints via company announcement, and approval decisions rest with regulators reviewing complete datasets.
Know your personal risk factors, protect your skin from UV, check your skin regularly, photograph anything you are watching so change is visible rather than remembered, and take new, changing, bleeding or non-healing lesions to a clinician.
Yes. Stage at diagnosis remains the dominant factor in melanoma outcomes. Better late-stage therapy does not reduce the value of finding a lesion earlier.