Applications of Natural Language Processing Tools in Orthopaedic Surgery: A Scoping Review

Abstract

The advent of many popular commercial forms of natural language processing tools has changed the way we can utilise digital technologies to tackle problems with big data. The objective of this review is to evaluate the current research and landscape of natural language processing tools and explore their potential use and impact in the field of orthopaedic surgery. In doing so, this review aims to answer the research question of how NLP tools can be utilised to streamline processes within orthopedic surgery. To do this, a scoping review was performed in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) and Arksey and O’Malley framework for scoping reviews, as well as a computer-assisted literature search on the Medline, Embase and Google Scholar databases. Papers that evaluated the use of natural language processing tools in the field of orthopaedic surgery were included. Our literature search identified 24 studies that were eligible for inclusion. Our scoping review captured articles that highlighted multiple uses of NLP tools in orthopaedics. In particular, one study reported on the use of NLP for intraoperative monitoring, six for detection of adverse events, five for establishing orthopaedic diagnoses, two for assessing the patient experience, two as an informative resource for patients, one for predicting readmission, one for triaging, five for auditing and one for billing and coding. All studies assessed these various uses of NLP through its tremendous computational ability in extracting structured and unstructured text from the medical record, including operative notes, pathology and imaging reports, and progress notes, for use in orthopaedic surgery. Our review demonstrates that natural language processing tools are becoming increasingly studied for use and integration within various processes of orthopaedic surgery. These AI tools offer tremendous promise in improving efficiency, auditing and streamlining tasks through their immense computational ability and versatility. Despite this, further research to optimise and adapt these tools within the clinical environment, as well as the development of evidence-based policies, guidelines and frameworks are required before their wider integration within orthopaedics can be considered.