Dear AuntMinnie Member,
Most of the discussion of artificial intelligence (AI) in radiology has swirled around the interpretation of medical images. But most astute observers believe that the first applications of AI will probably come in areas outside of image reading.
Take MRI contrast, for example. Deciding which patients should get contrast for their musculoskeletal MRI exams can be time-consuming. What if you could train a computer to do it?
A group from the University of California, San Francisco (UCSF) did just that in a new study we're highlighting in our Artificial Intelligence Community. The UCSF researchers used the IBM Watson algorithm to decide which patients were best suited for contrast before their MRI exams. The group then compared Watson's performance with that of a radiologist and more traditional machine-learning software.
Find out how well it worked by clicking here, or visit the community at ai.auntminnie.com.
Tracking CT radiation dose
Keeping track of radiation dose has been a thorny problem for imaging facilities. Fortunately, new tools are becoming available to make this easier -- to the benefit of patients.
One of these is radiation dose tracking software; it not only keeps track of the dose delivered to patients but can also be used to compare dose between different sites -- within a health network or across the globe -- when incorporated into a dose registry.
When researchers from Switzerland used dose tracking software and dose registries for CT scans, they found disturbing differences in the radiation dose being used at various sites around the world. Learn more by clicking here, or visit our CT Community at ct.auntminnie.com.
Radiology in Africa
We're pleased to bring you Dr. Sam Friedman's story of his recent trip to Africa on behalf of Rad-Aid International, a nonprofit group that works to improve the quality of medical imaging in developing countries. Sam just returned from Tanzania, and he describes his work training African residents and radiologists in PACS and nuclear medicine. Learn more by clicking here.
PACSman returns
Another familiar face is back in our Imaging Informatics Community: The PACSman, aka Michael J. Cannavo. From his home office in central Florida, Mike offers his thoughts on how shopping for a new PACS can be like riding out a hurricane in an article you can read by clicking here.
