HealthTech Mental Health Enterprise AI Quantum About Contact
Offering 01 — HealthTech

The clinic fits in your pocket.

Smartphone cameras are now medical-grade sensors. We've built deep-vision AI that turns a photo of skin, eye or smile into clinically actionable insight — validated against dermatological, ophthalmological and dental ground truth.

DEEPROOT · SKIN
Analysis complete · 2.1s
ABCDE score Asymmetric · border irregular
Confidence 87.4%
Recommendation Dermatologist referral
A — Product SuiteWhat we ship

Four instruments, one pocket.

border asymmetry evolution
01 · Dermatology

Skin cancer triage

Camera-based AI that analyses moles and lesions using the ABCDE framework — Asymmetry, Border, Colour, Diameter, Evolution — and flags suspicious patterns for dermatologist referral. Trained on over 200,000 dermoscopy and smartphone images across Fitzpatrick skin types I–VI.

Sensitivity94.1%
Specificity89.7%
Inference< 2s on-device
iOS · Android On-device ML CE marking in progress
MILD SEVERE
02 · Dermatology

Acne grading & care pathway

Image-driven acne assessment using the IGA (Investigator's Global Assessment) scale. Counts inflammatory lesions, comedones and nodules, grades severity, and generates personalised routines — or routes the patient to a dermatologist consultation within the app.

IGA concordance91%
Lesion types7 classes
Progress trackingWeekly
Teledermatology Progress photos Skincare routines
03 · Ophthalmology

Retinal screening

A retinal-image AI that screens for diabetic retinopathy, hypertensive retinopathy and early glaucoma indicators — using fundus photography captured via attachable smartphone optics. Clinically calibrated to the ETDRS scale.

DR sensitivity96.3%
Grades5 (ETDRS)
Referral triageAutomated
Primary care Rural outreach DICOM export
B — PipelineFrom photo to pathway

Four steps to clinical-grade insight.

01

Capture

A smartphone photo under standardised guidance — auto-exposure, glare detection, distance calibration.

02

Analyse

On-device vision models segment the region, extract features, and score against a medical reference ontology.

03

Ground

Every output traces to a verifiable source: clinical guideline, peer-reviewed study, or ontology entry.

04

Connect

When thresholds are crossed, the patient is routed — in the same app — to a qualified specialist for teleconsultation.

94%
Clinical agreement on dermoscopy cases
<2s
Median on-device inference latency
IVI
Fitzpatrick skin types in training data
200k+
Clinician-labelled images
C — ComplianceBuilt for regulation

Regulation is our starting
point, not our hurdle.

Clinical AI lives or dies on trust. Every Deeproot HealthTech product is built from day one to satisfy the toughest regulatory frameworks — because retrofit compliance is a contradiction in terms.

UK GDPR

Patient data stays in-region, encrypted at rest and in transit.

MHRA pathway

Software as Medical Device (SaMD) classification and registration planned.

ISO 13485

Medical device quality management system in development.

HIPAA-ready

US deployment posture with BAA coverage and audit trails.

ISO 27001

Enterprise-grade information security management.

Explainable AI

Every prediction attaches a reasoning trace the clinician can interrogate.

Inaugural partner

MediQSol / AI-native dental platform

Our inaugural HealthTech partner is redefining what a dental practice can be. A single secure platform unifies clinical staff, patients, partners and dentists — with AI embedded at every layer, from X-ray interpretation to treatment planning to appointment logistics. Ontology-grounded. Clinician-supervised. Patient-centred.

4
Stakeholder modules
100%
AI-native from day one
ISO 27001
Security posture
Become a partner
Clinical intelligence that earns its place in the workflow

Your patients deserve truth, not confidence.