Dear Digital X-Ray Insider,
In this edition of the Digital X-Ray Insider, we describe the efforts of a research group from Duke University in Durham, NC, to build a better computer-aided detection algorithm…and possibly help digital radiography gain some ground lost to CT in lung cancer imaging.
The group's idea was to start with the principles of tomosynthesis-based imaging, in which multiple 2D projections are taken at different angles to create a single 3D image. If CAD was good for a single 2D projection image, it ought to be even better when applied to multiple tomosynthesis images of the same area, the thinking goes.
The group found that the 3D CAD analysis, which they called correlation imaging, was able to reduce the number of false positives in phantom images. But the algorithm's results were less powerful with the human subjects that were part of the study. Find out more about the results by reading our Insider Exclusive, which you can reach by clicking here.
In other digital x-ray news, our recent articles on dose creep when converting to DR and CR have been generating strong reader interest. We revisit the topic with an article by consultant Robert Meisch, who explains how to vanquish the "dark side" of digital x-ray. Get more by clicking here.
For the rest of the news in the world of digital x-ray, visit the Digital X-Ray Community at xray.auntminnie.com, and feel free to send any ideas, comments, and suggestions for future articles on digital x-ray to me at [email protected].
