From NYU Langone radiologist and data scientist (Laura Heacock and Krzyzstof Geras.
We are in the early stages of planning a radiology eye-tracking project looking at ways to improve teaching radiology skills.
We were currently considering using eye tracking to test improvement in visual perception of areas of interest (suspicious abnormalities) on mammograms or breast ultrasounds before and after a user completes looking at a set of radiology images. The eye tracking would only be needed for the pre- and post-test. We have the images and the areas of concern, but do not have experience with eye tracking.
Later versions of this project would incorporate our AI radiology algorithms, but we’d need to have this data first in order to plan these further approaches.
Right now I was looking for someone interested in writing it up as a paper. If it’s successful as a pilot, there are several radiology grants that might able to fund future work.
So possibly more targeted to students at this point. We have the expertise in radiology and some prelim work to date on AI vs human perception that is a good starting point.
Contact: Laura Heacock laura@tethys.ai