{"title":"Analyzing Racial Differences in Imaging Joint Replacement Registries Using Generative Artificial Intelligence: Advancing Orthopaedic Data Equity","authors":"","doi":"10.1016/j.artd.2024.101503","DOIUrl":null,"url":null,"abstract":"<div><h3>Background</h3><div>Discrepancies in medical data sets can perpetuate bias, especially when training deep learning models, potentially leading to biased outcomes in clinical applications. Understanding these biases is crucial for the development of equitable healthcare technologies. This study employs generative deep learning technology to explore and understand radiographic differences based on race among patients undergoing total hip arthroplasty.</div></div><div><h3>Methods</h3><div>Utilizing a large institutional registry, we retrospectively analyzed pelvic radiographs from total hip arthroplasty patients, characterized by demographics and image features. Denoising diffusion probabilistic models generated radiographs conditioned on demographic and imaging characteristics. Fréchet Inception Distance assessed the generated image quality, showing the diversity and realism of the generated images. Sixty transition videos were generated that showed transforming White pelvises to their closest African American counterparts and vice versa while controlling for patients’ sex, age, and body mass index. Two expert surgeons and 2 radiologists carefully studied these videos to understand the systematic differences that are present in the 2 races’ radiographs.</div></div><div><h3>Results</h3><div>Our data set included 480,407 pelvic radiographs, with a predominance of White patients over African Americans. The generative denoising diffusion probabilistic model created high-quality images and reached an Fréchet Inception Distance of 6.8. Experts identified 6 characteristics differentiating races, including interacetabular distance, osteoarthritis degree, obturator foramina shape, femoral neck-shaft angle, pelvic ring shape, and femoral cortical thickness.</div></div><div><h3>Conclusions</h3><div>This study demonstrates the potential of generative models for understanding disparities in medical imaging data sets. By visualizing race-based differences, this method aids in identifying bias in downstream tasks, fostering the development of fairer healthcare practices.</div></div>","PeriodicalId":37940,"journal":{"name":"Arthroplasty Today","volume":null,"pages":null},"PeriodicalIF":1.5000,"publicationDate":"2024-09-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.sciencedirect.com/science/article/pii/S2352344124001882/pdfft?md5=4b249b4a85fd37c1ec3f9ccffbaea7c1&pid=1-s2.0-S2352344124001882-main.pdf","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Arthroplasty Today","FirstCategoryId":"1085","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S2352344124001882","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q3","JCRName":"ORTHOPEDICS","Score":null,"Total":0}
引用次数: 0
Abstract
Background
Discrepancies in medical data sets can perpetuate bias, especially when training deep learning models, potentially leading to biased outcomes in clinical applications. Understanding these biases is crucial for the development of equitable healthcare technologies. This study employs generative deep learning technology to explore and understand radiographic differences based on race among patients undergoing total hip arthroplasty.
Methods
Utilizing a large institutional registry, we retrospectively analyzed pelvic radiographs from total hip arthroplasty patients, characterized by demographics and image features. Denoising diffusion probabilistic models generated radiographs conditioned on demographic and imaging characteristics. Fréchet Inception Distance assessed the generated image quality, showing the diversity and realism of the generated images. Sixty transition videos were generated that showed transforming White pelvises to their closest African American counterparts and vice versa while controlling for patients’ sex, age, and body mass index. Two expert surgeons and 2 radiologists carefully studied these videos to understand the systematic differences that are present in the 2 races’ radiographs.
Results
Our data set included 480,407 pelvic radiographs, with a predominance of White patients over African Americans. The generative denoising diffusion probabilistic model created high-quality images and reached an Fréchet Inception Distance of 6.8. Experts identified 6 characteristics differentiating races, including interacetabular distance, osteoarthritis degree, obturator foramina shape, femoral neck-shaft angle, pelvic ring shape, and femoral cortical thickness.
Conclusions
This study demonstrates the potential of generative models for understanding disparities in medical imaging data sets. By visualizing race-based differences, this method aids in identifying bias in downstream tasks, fostering the development of fairer healthcare practices.
期刊介绍:
Arthroplasty Today is a companion journal to the Journal of Arthroplasty. The journal Arthroplasty Today brings together the clinical and scientific foundations for joint replacement of the hip and knee in an open-access, online format. Arthroplasty Today solicits manuscripts of the highest quality from all areas of scientific endeavor that relate to joint replacement or the treatment of its complications, including those dealing with patient outcomes, economic and policy issues, prosthetic design, biomechanics, biomaterials, and biologic response to arthroplasty. The journal focuses on case reports. It is the purpose of Arthroplasty Today to present material to practicing orthopaedic surgeons that will keep them abreast of developments in the field, prove useful in the care of patients, and aid in understanding the scientific foundation of this subspecialty area of joint replacement. The international members of the Editorial Board provide a worldwide perspective for the journal''s area of interest. Their participation ensures that each issue of Arthroplasty Today provides the reader with timely, peer-reviewed articles of the highest quality.