We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
Artificial intelligence (AI) and large data registries: Understanding the advantages and limitations of contemporary data sets for use in AI research.
- Authors
Kunze, Kyle N.; Williams, Riley J.; Ranawat, Anil S.; Pearle, Andrew D.; Kelly, Bryan T.; Karlsson, Jon; Martin, R. Kyle; Pareek, Ayoosh
- Abstract
This article explores the benefits and limitations of using large data registries for artificial intelligence (AI) research in orthopaedics. It discusses two main types of data sources: institutional databases, which provide detailed patient information but require maintenance, and administrative databases, which have a larger pool of data but may lack specific details. While large data registries offer advantages such as robust sample sizes and readily available data, they also have limitations including coding inconsistencies and missing critical information. The article emphasizes the need for AI-specific data registries that capture diverse patient data to improve the accuracy and applicability of AI models in orthopaedic research and patient care.
- Subjects
ARTIFICIAL intelligence; JOINT infections; KNEE pain; MACHINE learning; TOTAL shoulder replacement; SURGICAL complications; ARTHROSCOPY; NATURAL language processing; ANTERIOR cruciate ligament surgery
- Publication
Knee Surgery, Sports Traumatology, Arthroscopy, 2024, Vol 32, Issue 1, p13
- ISSN
0942-2056
- Publication type
Article
- DOI
10.1002/ksa.12018