Artificial intelligence in the management and treatment of burns: a systematic review

Abstract

Background: Artificial intelligence (AI) is an innovative field with potential for improving burn care. This article provides an updated review on machine learning in burn care and discusses future challenges and the role of healthcare professionals in the successful implementation of AI technologies.

Methods: A systematic search was carried out on MEDLINE, Embase and PubMed databases for English-language articles studying machine learning in burns. Articles were reviewed quantitatively and qualitatively for clinical applications, key features, algorithms, outcomes and validation methods.

Results: A total of 46 observational studies were included for review. Assessment of burn depth (n = 26), support vector machines (n = 19) and 10-fold cross-validation (n = 11) were the most common application, algorithm and validation tool used, respectively.

Conclusion: AI should be incorporated into clinical practice as an adjunct to the experienced burns provider once direct comparative analysis to current gold standards outlining its benefits and risks have been studied. Future considerations must include the development of a burn-specific common framework. Authors should use common validation tools to allow for effective comparisons. Level I/II evidence is required to produce robust proof about clinical and economic impacts.

Keywords: Artificial intelligence; Burn; Computer vision; Machine learning; Neural networks.

Figure 1.

Trend of artificial intelligence burns-related publications on PubMed

Figure 2.

Domains of artificial intelligence. Physical artificial intelligence relates to machines interacting with their physical environment whereas virtual artificial intelligence is represented by machine learning

Figure 3.

Illustration of different subtypes of machine learning. (a) Supervised learning involves the feeding of labelled data allowing the computer to create a predictive algorithm of a known output to correctly classify the depth of the burns. (b) Unsupervised learning uncovers any patterns such as the categorisation of burns depth from the unlabelled data. (c) Reinforcement learning is the process to successfully match the input and output data while learning from its successes and failures. It may share features of both a supervised and unsupervised process

Figure 4.

Representation of artificial neural networks. Artificial neural networks can independently process signals in layers of simple computational units. At the input level neurons receive information, perform a calculation and transmit output to the next neurone in the hidden layer. Within the hidden layers, calculations are carried out to analyse and extract the complex patterns in the dataset. The data is then passed onto the output layer that provides the final step in the analysis for interpretation. Deep learning involves the learning of more complex and subtle patterns than a simple one- or two-layer neural network

Figure 5.

Flowchart showing systematic literature attrition

Figure 6.

Number of machine learning articles on the different applications on burn patient care. ML machine learning

Figure 7.

Algorithms used in burn-related machine learning articles

Figure 8.

Methods of model validation used in burn-related machine learning articles.CV cross-validation

Figure 9.

Possible benefits of artificial intelligence at the different stages of a patient with a burn injury. AKI acute kidney injury, TBSA total body surface area