the surgical image registration generator (sirgn) baseline

Department: Computer Science & Engineering
Faculty Advisor(s): Ryan Kastner

Primary Student
Name: Michael Joseph Barrow
Phone: 858-534-8908
Grad Year: 2019

Augmented reality (AR) is a promising technique for enhancing surgical procedures by positioning image guidance aids on video of the surgical scene. Non-rigid 3D registration is an essential step in surgical AR; it is used to determine the position of organ surfaces so image guidance aids can be projected on the surgical scene correctly. We introduce SIRGn -- a novel dense, granular baseline quality metric for video registration algorithms. Our method provides a triangular mesh overlay representing the quality of registration and can highlight areas of unacceptably poor registration. We evaluate the metric by comparing two surface registrations of surgical video

Industry Application Area(s)
Life Sciences/Medical Devices & Instruments

« Back to Posters or Search Results