The Burn Medical Assistant: Developing Machine Learning Algorithms to Aid in the Estimation of Burn Wound Size
Abstract
We will test the hypothesis that a computer machine learning algorithm can analyze and classify burn injures using multispectral imaging within 5% of an expert clinician through a series of steps. First, we will develop a model to map images of burn wounds captured from a portable multispectral camera to a standard Lund-Browder burn wound diagram. We will then have burn wound experts estimate the burn wound size using an enhanced burn wound diagram. The mappings made by burn wound experts will then be compared to automated measurements. Finally, through the use of machine learning (ML), we will build models from these multispectral images and expert burn wound mapping to create a predictive model of burn wound size and depth that will approximate an expert's opinion to within 5%.
Document Details
- Document Type
- Technical Report
- Publication Date
- Oct 01, 2017
- Accession Number
- AD1048286
Entities
People
- Christopher Argenta
- Gregory Rule
- Jeremy Pamplin
- Maria Serio-melvin