A practical guide to the development and deployment of deep learning models for the orthopedic surgeon: part II

Abstract

Deep learning has the potential to be one of the most transformative technologies to impact orthopedic surgery. Substantial innovation in this area has occurred over the past 5 years, but clinically meaningful advancements remain limited by a disconnect between clinical and technical experts. That is, it is likely that few orthopedic surgeons possess both the clinical knowledge necessary to identify orthopedic problems, and the technical knowledge needed to implement deep learning‐based solutions. To maximize the utilization of rapidly advancing technologies derived from deep learning models, orthopedic surgeons should understand the steps needed to design, organize, implement, and evaluate a deep learning project and its workflow. Equipping surgeons with this knowledge is the objective of this three‐part editorial review. Part I described the processes involved in defining the problem, team building, data acquisition, curation, labeling, and establishing the ground truth. Building on that, this review (Part II) provides guidance on pre‐processing and augmenting the data, making use of open‐source libraries/toolkits, and selecting the required hardware to implement the pipeline. Special considerations regarding model training and evaluation unique to deep learning models relative to “shallow” machine learning models are also reviewed. Finally, guidance pertaining to the clinical deployment of deep learning models in the real world is provided. As in Part I, the focus is on applications of deep learning for computer vision and imaging.

A practical guide to the development and deployment of deep learning models for the orthopaedic surgeon: Part III, focus on registry creation, diagnosis, and data privacy

“…This extracted, structured data can be used to build and augment orthopaedic data registries, which may serve as cornerstones of orthopaedic research. Below, an illustration of how the deep learning-based concepts discussed in Parts I and II [25,26], such as image classification and segmentation, have been applied to augment orthopaedic registries is provided. Specifically, these are discussed in the context of (1) image classification, (2) image segmentation and (3) NLP-based medical report analysis.…”

“…Deep learning, and specifically computer vision, has a similar potential, although a greater understanding of the steps needed to leverage its potential by orthopaedic surgeons with clinical expertise and knowledge of the problems it can solve may be needed. As a brief review from Parts I and II of this series, deep learning algorithms commonly focus on image-based tasks and involve training neural networks with multiple layers to learn patterns from data in a hierarchical fashion, while NLP algorithms are specifically designed to analyse and understand human language [25,26].…”

“…In Parts I and II of this series, a high-level overview of the process of leading a computer vision-based project from the orthopaedic surgeon perspective was provided, including all steps from defining the problem to model deployment in the clinical setting [25,26]. In Part III, practical examples of how the strategies discussed in the prior two articles can be used to transform multiple areas of orthopaedic surgery are provided: reduction of administrative burden, prevention of complications, preoperative planning and others.…”

Enabling Personalized Medicine in Orthopaedic Surgery Through Artificial Intelligence

“…Setting up Skills: Who and How? Successful institutional AI process development and implementation requires a multidisciplinary team of individuals with clinical, imaging, and technology/analytic backgrounds 10,11 (Fig. 2).…”

“…AI is a term that is frequently used and often poorly understood 5 . Figure 1 provides a list of terms commonly used throughout this article and their definitions 4,[6][7][8][9][10][11] . AI refers to any computer system that is capable of problem solving and learning from experience, much like the way humans learn 4 .…”