Guideline-aware safeguards
Safety boundaries are designed around recognised clinical review expectations without publishing internal logic.
In clinical validation · UK design partners welcome
Oral lesion triage built around clinician review.
OralPath is clinician-in-the-loop AI triage support for oral mucosal lesions. It is designed to help clinical teams review concerning presentations earlier while keeping responsibility and judgement with the clinician.
Clinician-in-the-loop
Guideline-aware
Clinician sign-off
Screens from the live OralPath build. Synthetic sample cohort only — no patient-identifiable information.
Features
Built for safer oral lesion review.
OralPath is designed as a clinician-controlled product experience, not a public specification. The page describes the clinical value and safety posture without exposing implementation detail.
Consistent clinical review
The product experience is designed to reduce variation in how relevant context is gathered and reviewed.
Clear review support
The interface is designed to make the support visible, challengeable and secondary to clinician judgement.
Clinician sign-off
Outputs remain under clinician control, with review and sign-off retained in the clinical workflow.
How it works
From case review to clinician sign-off.
The public workflow is deliberately high-level. Detailed product behaviour is not published on the open web.
-
01
Review
The clinician reviews the presenting context inside a guided product experience.
-
02
Support
OralPath provides decision-support context without replacing clinical judgement.
-
03
Safeguard
Guideline-aware safeguards sit around the product experience.
-
04
Govern
Clinical safety, auditability and deployment evidence are part of the programme.
-
05
Sign
The responsible clinician remains accountable for the final decision.
Designed for triage
Product-market fit without exposing the product map.
OralPath is positioned for clinical triage environments where oral mucosal lesion review needs better structure, safety and governance. Detailed workflow design is shared only with appropriate partners.
Pathway logic
Clinician controlled
Image
Clinical material reviewed in context
Context
Relevant information held together
Safety
Guideline-aware safeguards
Decision
Clinician review and sign-off
Public preview intentionally limited
Safety by design
Clinical judgement stays in charge.
Triage support only
OralPath supports triage review. It does not diagnose oral cancer.
Safety safeguards
Safety boundaries are part of the product programme, with details reserved for partner contexts.
Human accountability
The responsible clinician reviews the full context and signs the referral decision.
Evidence before claims
OralPath is built and undergoing clinical validation; safety and performance evidence is published as it reads out (2026/27).
FAQ
Questions from clinical teams.
OralPath is available for guided demos with appropriate NHS and dental teams. Public materials intentionally avoid product implementation detail.
What is OralPath?+
OralPath is clinician-in-the-loop AI triage support for oral mucosal lesions. It is built to support safer clinical review without replacing clinician judgement.
Does OralPath diagnose oral cancer?+
No. It is not a diagnosis and does not replace the clinician. It surfaces risk and pathway support for review and sign-off.
Do you publish the product logic?+
No. Public materials describe positioning, safety boundaries and clinical context only. Detailed workflow and implementation details are shared only in appropriate demo or partner contexts.
Who is OralPath for?+
It is being developed for clinical teams where oral mucosal lesion review needs better structure, safety and auditability.
How can we see it?+
Request a guided demo through the booking page or email cs@csong.health with your organisation and role. Please do not send patient-identifiable data.
Design partners
Review OralPath in a controlled demo.
Appropriate clinical teams can review the product in context without public disclosure of product logic or implementation detail.