Pressure Ulcer Management System Using Machine Learning

  • Research type

    Research Study

  • Full title

    Pressure Ulcer Prevention Using Machine Learning: An Automated Skin Damage and Pressure Ulcer Assessment Tool for Nursing Professionals, Patients, Family Members and Carers

  • IRAS ID

    253949

  • Contact name

    Paul Fergus

  • Contact email

    p.fergus@ljmu.ac.uk

  • Sponsor organisation

    Mersey Care NHS Foundation Trust

  • Clinicaltrials.gov Identifier

    NA, NA

  • Duration of Study in the UK

    0 years, 7 months, 29 days

  • Research summary

    This is an initial pilot study to determine the accuracy of a Pressure Ulcer Management system to categorise and report the characteristics associated with each category of a pressure ulcer. The system has been developed using the state-of-the-art in Machine Learning and Object Detection to initially train our system using 700 images obtained from Google Images. There are currently seven categories the system can report on: category I, II, III and IV, and Eschar, Slough and Granulating. In this initial study we would like to collect high quality photographs of pressure ulcers from patients and using our system automatically categorise the wound and report associated characteristics to obtain baseline results. To evaluate the efficacy of the system specialist nurses within the skin service will report on the decisions made by the machine learning system. This will provide two initial outcomes. The collection of high quality pressure ulcer images for future system training and clinical feedback on the quality of automated pressure ulcer classification and reporting and its potential use as a tool in clinical practice. If successful the pressure ulcer tool would help to standardise treatment and reporting, aid education, and act as a support tool for medical practitioners that encounter or treat pressure ulcers in day-to-day care.

  • REC name

    East Midlands - Leicester South Research Ethics Committee

  • REC reference

    19/EM/0065

  • Date of REC Opinion

    21 Mar 2019

  • REC opinion

    Further Information Favourable Opinion