Assessing deep learning accuracy in the measurement of radiographic parameters in pediatric hip X-rays

Byoung-Dai Lee, Jin Young Kim, Ki-Ryum Moon, Mu Sook Lee · 2025

Read Assessing deep learning accuracy in the measurement of radiographic parameters in pediatric hip X-rays by Byoung-Dai Lee, Jin Young Kim, Ki-Ryum Moon, Mu Sook Lee (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/pmc-12750828.txt.

Description

Checking how a child's hip is developing means measuring several angles and indices on pelvic X-rays, slow work that varies between readers. Byoung-Dai Lee and three co-authors built a deep learning system that measures four of these quantities automatically: the acetabular index, the Shenton line, and indices of pelvic rotation and tilt. It was trained on 1,495 X-rays of healthy Korean children and tested on a separate set of 1,300.

The system's measures closely matched those of radiologists, with agreement scores between 0.89 and 0.97 across the four quantities and small average errors. Accuracy shifted somewhat with age, reflecting how the pelvic bones change as children grow. The authors conclude that the system is reliable enough to help separate normal hips from dislocated ones and to judge whether a pelvic X-ray is well positioned.

Catalog record

CREATORByoung-Dai Lee, Jin Young Kim, Ki-Ryum Moon, Mu Sook Lee
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE37 KB
VERSION1.0
LICENSECC BY 4.0
RIGHTSCC BY
SOURCEEurope PMC

Source: https://europepmc.org/articles/PMC12750828

Browse the full library · About the Tower of Babel