XRAIconnect Image Guidance Software for ERCP Procedures

Patients scheduled for an ERCP procedure will undergo pre-operative MRCP imaging scans (MRI scans) which will be reconstructed into digital, 3D models of the biliary and pancreatic ducts and surrounding organs/pathology.

At the ERCP visit and using XRAIconnect software, NYULH ERCP operators will align these models with live, intra-operative 2D X-ray (fluoroscopy) videos/images (registration), allowing operators to view anatomical structures in 3D, with (stereoscopically) or without the use of an extended reality head-mounted display. The accuracy of XRAIconnect in terms of mean fiducial registration error and mean multiplanar triangulation error will be measured.

Medical device software that generates three-dimensional (3D) geometric models from pre-operative imaging, and is capable of aligning them to live 2D fluoroscopic imaging during interventions, to provide enhanced navigational cues and a native 3D understanding of the spatial relationships between instrumentation and patient anatomy. In the context of ERCP, XRAIconnect uses data from pre-operative MRCP data to generate 3D models of the hepatobiliary system and surrounding structures which can be aligned with live, intra-operative 2D X-ray (fluoroscopy) videos/images during ERCP in a process known as registration.

Mean fiducial registration error is a measure of goodness of fit of the registration; measured in mm.

Mean multiplanar triangulation error is the error due to small inconsistencies in registration across the imaging planes which may cause the resulting 3D rays cast not to intersect at a single, consistent point in space, leading to a slight mismatch between the 2D and 3D correspondences; measured in mm.