Development of an Explainable Artificial Intelligence Model for Asian Vascular Wound Images
Care Process & Redesign
Technology
Singapore Health Biomedical Congress
National Healthcare Group
31 December 2023
To develop an explainable AI model for vascular wound image analysis in an Asian population. Model supports wound classification, measurement, and segmentation, enhancing wound care in diverse settings.
Year Submitted: 2023
Published Date: 31 December 2023
Tags: Technology, Care Process ＆ Redesign, Digital Health, Digitalisation, Automation, Artificial Intelligence, Product Development, Proof Of Concept
About this Content
Aims
To develop an explainable AI model for vascular wound image analysis in an Asian population.
Background
Chronic wounds require consistent and precise assessment; AI offers a scalable solution.
Methods
Trained AI on 2,957 vascular wound images, developed explainable web app to evaluate wound features.
Results
Model achieved AUROC of 0.99 for classification and explainability of 60.32%; application integrates with clinical workflows.
Conclusion
Model supports wound classification, measurement, and segmentation, enhancing wound care in diverse settings.
Lessons Learnt
Explainable AI enhances trust and integration of AI in clinical wound care.
Additional Information
Singapore Health & Biomedical Congress 2023: Silver Award
Keywords
Artificial Intelligence, Wound Analysis
Innovators' Details
Innovators' Details
Healthcare Cluster(s) | National Healthcare Group |
Organization(s) Involved | Woodlands Health, Tan Tock Seng Hospital, Nanyang Technological University |
Platform(s) | Singapore Health Biomedical Congress |
Healthcare Professional Group(s) | Nursing, Medical, Healthcare Administration |
Applicable Specialty or Discipline | Surgery |
Project Lead(s) | Zhiwen Joseph Lo |
Project Member(s) | Malcolm Han Wen Mak |
Connect with this contributor!
Ms Liang Shanying - shanying_liang@wh.com.sg