Communications of the ACM
Researchers Create AI-Powered Deep Learning Model to Support Medical Diagnostics
A deep learning model can learn to identify diseases from medical scans faster and more accurately, according to new research.
Credit: Shutterstock
Computer scientists at Canada's University of Alberta (U of A) and MEDO, a U of A spinoff company, developed a deep learning model capable of identifying diseases from medical scans quickly and more accurately.
The researchers aimed to address a major challenge associated with medical diagnostics: privacy issues that prevent researchers from accessing more than a few hundred medical scans.
Their algorithm was trained on medical images and their corresponding "probabilistic" diagnosis.
U of A's Roberto Vega said, "Our approach both improved the classification accuracy of the model and provided a meaningful confidence in its prediction, giving an estimate of the probability that a disease is present in a scan."
From Folio (University of Alberta, Canada)
View Full Article
Abstracts Copyright © 2021 SmithBucklin, Washington, DC, USA
No entries found