Investigators at University of Basel Describe Findings in Cardiovascular Research (Reduction In Radiologist Interpretation Time of Serial Ct and Mr Imaging Findings With Deep Learning Identification of Relevant Priors, Series and Finding...).

A recent study conducted at the University of Basel in Switzerland has found that a deep learning tool can significantly reduce the time and user interactions required to assess findings of interest on cross-sectional imaging with relevant prior exams. The tool, called TimeLens (TL), uses natural language processing and descriptor-based image-matching algorithms to automatically identify and display relevant prior studies. The study involved nine radiologists who performed two reading sessions, one without TL and one with TL, and found that TL reduced the average time needed to assess a finding by 40.1%. The tool also reduced the number of mouse clicks and mouse distance traveled. The research concluded that TL has the potential to improve the efficiency of radiology workflows. [Extracted from the article]