Leveraging advanced facial recognition technologies to annotate a large corpus of patient photos

Poster Number: P129
Presentation Time: 05:00 PM - 06:30 PM

Abstract Keywords: Evaluation, Healthcare Quality, Real-World Evidence Generation
Primary Track: Foundations
Programmatic Theme: Consumer Health Informatics

We investigated the use of Microsoft's Face AI API to annotate a large corpus of patient photos with the following attributes: size, position, center, angle and occlusions of the face, exposure and blur level of the photo, the complexity of background and number of faces in one photo. We present the detailed steps needed to secure the annotation environment and the evaluation results of a test set of photos leveraging this annotation tool.

Speaker(s):
Chenyang Li, MD
Columbia University in the City of New York