MedTech AI Wound Assessment Platform — UK Community Nursing

AI Wound Classification Model: Training dataset: 42,000 annotated wound images (pressure ulcers, diabetic foot ulcers, surgical wounds, venous leg ulcers) — licensed from Tissue Analytics (US) + 8,000 NHS community nursing images under HRA ethics. Model: MobileNetV3 (optimised for mobile inference — runs on iPhone 12+/Android equivalent). Output: wound type (5 classes), healing trajectory (improving/static/deteriorating), infection probability (0–1), and area estimation from camera geometry. Inference: on-device (CoreML on iOS, TensorFlow Lite on Android) — no cloud round-trip required for clinical assessment. MHRA Class IIa SaMD Classification: MHRA SaMD classification: wound assessment tool that informs clinical decisions (not replaces) = Class IIa. Intended purpose limitation: AI provides wound classification for information — nurse makes clinical decision. MHRA Technical File: classification rationale, clinical evaluation report (systematic literature review of wound assessment AI performance: sensitivity/specificity for pressure ulcer detection), and post-market surveillance plan (6-month PMCF study). SNOMED CT Integration for Community Nursing: NHS SNOMED CT coding: wound assessment outcomes automatically coded to SNOMED CT wound care hierarchy. Community nursing system (SystmOne/EMIS Web) integration: FHIR R4 Observation resource with SNOMED CT code for wound type, healing status, and intervention. Automated clinical outcome data: wound healing trajectory tracked per patient — dashboard showing community nursing team wound care outcomes vs NHSE benchmarks. Mobile-First Clinical Workflow: React Native cross-platform (iOS + Android — NHS community nurses use both). Offline-first: complete wound assessment without network connectivity (common in community settings — homes, care homes without WiFi). Background sync: assessments queued locally → synced to server when connectivity restored. Camera API: guided photograph (alignment guide overlay, consistent lighting check, scale reference detection). WCAG 2.1 AA: mobile accessibility — voice-over testing on iOS for nurses with visual impairment.