An AI-powered software system has been trained to mimic radiologists’ eye movements as well as predict where a radiologist would likely look next, according to an April 21 in Scientific Reports.
Here are four things to know about the technology:
- The system, called MedGaze, was developed by researchers from the University of Houston and the University of Texas MD Anderson Cancer Center, also in Houston, according to a May 20 UH news release.
- MedGaze was trained on thousands of eye-tracking sessions recorded while radiologists interpreted chest X-rays, imaging and radiology report datasets.
- For the study, researchers evaluated 20 of the system’s predicted scanpaths. Thirteen of the scanpaths “closely resembled expert patterns,” and 18 covered 60% to 80% of the key regions scanned by experts.
- “MedGaze’s ability to minimize redundancy and emulate expert gaze behavior enhances training and diagnostics, offering valuable insights into radiologist decision-making and improving clinical outcomes,” the study authors wrote.
Read the full study here.
