A multi-scene deep learning model for automated segmentation of acute vertebral compression fractures from radiographs: a multicenter cohort study

Hao Zhang, Genji Yuan, Ziyue Zhang, Xiang Guo, Ruixiang Xu, Tongshuai Xu, Xin Zhong, Meng Kong, Kai Zhu, Xuexiao Ma · 2024

Read A multi-scene deep learning model for automated segmentation of acute vertebral compression fractures from radiographs: a multicenter cohort study by Hao Zhang, Genji Yuan, Ziyue Zhang, Xiang Guo, Ruixiang Xu, Tongshuai Xu, Xin Zhong, Meng Kong, Kai Zhu, Xuexiao Ma (2024) 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-11612108.txt.

Description

Fresh fractures of the spine from crushed vertebrae need fast, exact outlining on X-rays before surgery. A ten-author team built a deep learning model, PFNet, that traces these fractures automatically and works across the varied images of real clinics. They trained it on radiographs from two hospitals and tested it on held-out patients plus three outside hospitals, over two thousand cases in all.

The model scored between about 98.5 and 100 percent accuracy across the test sets, beating earlier approaches on every measure the team compared. Heat maps showed the model attends to the fractured regions a human reader would inspect, which supports trusting its outlines. The authors present the tool as the first of its kind built for the varied scenes of everyday practice, and as ready support for planning and guiding repair operations.

Catalog record

CREATORHao Zhang, Genji Yuan, Ziyue Zhang, Xiang Guo, Ruixiang Xu, Tongshuai Xu, Xin Zhong, Meng Kong, Kai Zhu, Xuexiao Ma
YEAR2024
COLLECTIONPAPERS
FORMATTXT
FILE SIZE30 KB
VERSION1.0
LICENSECC BY 4.0
RIGHTSCC BY
SOURCEEurope PMC

Source: https://europepmc.org/articles/PMC11612108

Browse the full library · About the Tower of Babel