Automatic Identification of Make and Model of Ankle Implants using Artificial Intelligence

Abstract

Orthopedic implant identification is a crucial step before planning a revision surgery. Failure to identify implants preoperatively can cause delay in surgeries. This increases pain and trauma to patients. Ankle replacement has seen an increase in both primary and revision surgeries recently. The paper proposes a novel framework to identify the make and model of the Ankle implants from X-ray images using Artificial intelligence. Authors have identified the implants with an accuracy of 96.09 % and an Area under curve of 0.9954 proving the superiority of deep learning in classifying the implants. The proposed work formulates a first and unique framework to identify the make and model of ankle replacements.

Ähnliche Arbeiten