But it’s far more efficient for artificial intelligence to scroll through those stacks of images. Then, the algorithm can spot ones with significant abnormalities and give them to radiologists for further examination.
Now, Geek.com reports researchers at UC Berkeley and San Francisco have made such AI.
“We wanted something that was practical, and for this technology to be useful clinically, the accuracy level needs to be close to perfect,” said Esther Yuh, Associate Professor of Radiology at University of California San Francisco.
Their final product took just one second to determine whether an entire head scan contained any signs of haemorrhage. It even found some small abnormalities that experts missed, and all with “an acceptable level” of false positives.
