Week in Review: Minnies finalists | AI for mammography | AI for CT

Dear AuntMinnie Member,

We were pleased this week to announce the finalists in the Minnies, our annual event recognizing excellence in radiology.

There are now two finalists in each of the 15 races, ranging from Most Influential Radiology Researcher to Best Radiology Image. Other Minnies categories cover devices, software, mobile apps, and more. Find out who made it to the next round at minnies.auntminnie.com.

Our expert panel is currently busy voting on the finalists, with winners to be announced in late October. Stay tuned!

AI for mammography

Could the use of artificial intelligence (AI) for mammography make diagnoses more accurate and also cut the workload of the average breast center? That's the suggestion of a new study we're highlighting from this week's Conference on Machine Intelligence in Medical Imaging (C-MIMI), sponsored by the Society for Imaging Informatics in Medicine (SIIM).

Researchers from Johns Hopkins University tested a deep-learning algorithm called DeepCAT (computer-assisted triage) to analyze a set of screening mammograms and remove studies that the algorithm classified as normal. The idea is that this AI application would enable radiologists to focus on the studies most likely to have malignancies. Find out how well it performed by visiting our Artificial Intelligence Community.

In other news from C-MIMI, Senior Editor Erik L. Ridley conducted video interviews with several imaging informatic luminaries at the show, including Paul Nagy, PhD, on the state of AI research; Dr. Bradley Erickson, PhD, on key trends in AI; and Dr. Eliot Siegel on the clinical adoption of AI.

AI for CT

Meanwhile, our coverage of another talk on AI this week by the peripatetic Dr. Siegel -- this time at the International Society for Computed Tomography (ISCT) -- was another of our most-read stories.

Dr. Siegel discussed how AI for CT could be the "killer app" that makes the technology truly achieve clinical success. But not for image interpretation: He believes that AI could help with CT image acquisition and data processing to deliver studies with improved image quality and lower radiation dose.

While you're in our CT Community, but sure to check out an article on cinematic rendering, a fascinating technology that produces amazingly lifelike images based on CT data. The technique could prove to be particularly useful for musculoskeletal imaging, according to researchers from Canada.

Also, learn how radiation dose from CT scans can be reduced when radiologists take the lead. These stories and more are available in the CT Community.

Page 1 of 99
Next Page