Deep Learning: Fracture Risk Assessment

• Approver: AHRA®
• ARRT®: Accepted
• Expiration Date: 6/30/25
• Credit Hours: TBD Category A Credit

Course Acceptance List

• ARRT: Accepted
• Florida: Cat A / Technical
• Texas: Direct
• NMTCB: Accepted

Objectives

• Understand the underutilization of Dual-energy X-ray absorptiometry (DXA) in measuring bone mineral density (BMD) and evaluating fracture risk and recognize the need for automated tools to enhance efficiency and accuracy in this domain.
• Explore the development and implementation of an automated tool for identifying fractures, predicting BMD, and evaluating fracture risk using plain radiographs, including the underlying algorithms and methodologies.
• Analyze the performance evaluation of the automated tool on large patient datasets, assessing its calibration, bias, and precision-recall characteristics for hip and lumbar spine BMD measurements, as well as its accuracy in predicting hip and spine osteoporosis, as well as high 10-year major and hip fracture risk.
• Gain practical skills in utilizing the automated tool effectively, including its application in identifying high-risk patients for osteoporosis and the interpretation of results with high positive or negative predictive value, comparing it to the conventional DXA method.

Course Description

This course explores the underutilization of Dual-energy X-ray absorptiometry (DXA) in measuring bone mineral density (BMD) and evaluating fracture risk. You will learn about an innovative automated tool designed to identify fractures, predict BMD, and assess fracture risk using plain radiographs. The course will examine the performance evaluation of the tool on a large dataset of patients, demonstrating its well-calibrated nature with minimal bias in hip and lumbar spine BMD measurements. You will also gain an understanding of the tool’s high accuracy and precision, with impressive results for identifying hip and spine osteoporosis as well as high 10-year major and hip fracture risk. By the end of the course, you will be equipped with the knowledge to utilize this automated tool effectively, enabling them to identify high-risk patients for osteoporosis with remarkable predictive value.