NURSE-AI: A Nurse-by-Design Framework for Multi-Sensor, AI-Enabled Chronic Wound Assessment in Community Healthcare

Area of Science:

• Medical Technology
• Artificial Intelligence
• Healthcare Informatics

Background:

• Chronic wound assessment in community healthcare faces challenges due to environmental variability and subjective evaluation, leading to measurement errors.
• Despite advancements in multi-sensor technologies and AI, real-world adoption is hindered by workflow misalignment, lack of interpretability, and regulatory hurdles.

Purpose of the Study:

• To present the NURSE-AI framework, a Nurse-by-Design methodological approach for evaluating and preparing multi-sensor, AI-enabled wound assessment systems for community healthcare.
• To assess the feasibility and usability of the NURSE-AI framework through a mixed-method evaluation with community nurses.

Main Methods:

• Development of a GDPR-compliant synthetic multimodal dataset (RGB, depth, thermal, hyperspectral-proxy).
• Workflow-embedded acquisition modeling tailored for Family and Community Nurses (FCNs).
• Integration of a Wound Bed Preparation (WBP)-aligned interpretability layer and a governance-by-design checklist for regulatory readiness.

Main Results:

• The System Usability Scale (SUS) score was 74.5 ± 6.2, indicating good usability.
• Synthetic data evaluation showed promising segmentation performance (IoU 0.87–0.93) and accurate wound area estimation (simulated ICC ≥ 0.90).
• Agreement between AI-generated WBP mappings and nurse interpretation was high (κ = 0.80–0.84).

Conclusions:

• The NURSE-AI framework offers a structured pathway for integrating sensor innovation, AI interpretability, and nursing workflows in community healthcare.
• This framework provides essential groundwork for future clinical validation and scalable deployment of AI-enabled wound assessment systems.