The new system currently includes FDA-cleared algorithms from five companies along with seven developed by Aidoc.
Experts cited positive findings but stopped short of recommending convolutional neural networks as the go-to standard of care for detecting ICH.
One possibility involves a "food label" style high-level summary of relevant data points, the college said in recently submitted comments.
More than 2,700 participants were tasked with developing AI models capable of completing two different tasks.
The algorithm doesn't add nor lose any image information needed to assess coronary plaques, experts said.
The tool can be universally applied across multiple CT scanners and vendors, experts reported in the European Journal of Radiology.
The software produced additional actionable findings and better case overviews, according to real-world, prospective data shared Wednesday.
The company's Insight CXR Triage tool can detect 10 common image anomalies with up to 99% accuracy.
Specifically, the deep learning platform distinguished minimally invasive adenocarcinomas from more deadly varieties.
In many instances, the assistive technology helped less-experienced providers change their image impressions and treatment decisions, experts explained.
The free tool just slightly overshot MRI volumes during the height of the pandemic, greatly outperforming manual estimates.
AI-assisted scans are six times faster than standard CMRIs and require fewer breath-holds, company representatives said Wednesday.
"Be the squeaky wheel. This issue was taken care of because physicians stood up,” Rad Partners' Richard Heller, MD, told RSNA attendees Monday.
The platform runs on Azure, a cloud computing service operated by Microsoft, which recently announced plans to acquire Nuance for $20B.
Radiology experts analyzed fetal MRIs from 33 women who contracted the virus while bearing a child, sharing their results Tuesday at RSNA.