Volumetric laser endomicroscopy (VLE) uses advanced optical coherence tomography for real-time, cross-sectional imaging to a depth of 3 mm and a resolution of 7 microns. The VLE probe, deployed through the endoscope’s working channel, scans 6 cm of esophagus in 90 seconds. While prior studies have demonstrated high sensitivity for VLE detection of dysplasia within Barrett’s esophagus (BE), there are cases where random tissue sampling confirms dysplastic BE despite an absence of concerning VLE findings. Our aim was to assess whether new software with Image Visualization Enhancement (IVE), which uses color to highlight specific VLE features including hyporeflective structures and loss of normal wall layering, can identify areas of concern missed on initial VLE review. A multicenter US registry prospectively collected data on 1000 patients at 18 sites when VLE was used for esophageal evaluation. All procedures used the NvisionVLE® Imaging System (NinePoint Medical, Bedford, MA), and all sites obtained Institutional Review Board approval prior to beginning enrollment. Data review identified patients found to have BE at least indefinite for dysplasia on any tissue sampling, after which all cases without a focally concerning VLE finding were selected (Fig 1). All VLE images for 16 of these cases were loaded onto a VLE console and presented to a gastroenterologist with expertise in BE assessment and VLE interpretation. Each case was reviewed as a standard VLE, after which IVE was turned on and a repeat evaluation was performed. Time to complete image review using each approach, as well as findings and an overall confidence rating for likelihood of dysplasia (1-5 scale) were documented. The mean review time for standard VLE was 4 minutes per scan (range 3-5). A mean of 2 (1-7) focal areas of concern for dysplasia were identified on each scan (Fig 2). Subsequent review with IVE required, on average, an additional 2 minutes (1-3); in every case, focal areas of concern on VLE corresponded to highlighted areas on IVE. In half of the cases (8/16), use of IVE identified at least one additional suspicious region not selected by the reviewer during standard VLE image review. The mean confidence level for dysplasia was 2.8 (1-5), indicating a moderate degree of certainty. While focal areas suspicious for dysplasia were found in all scans deemed not concerning by the original endoscopist, using IVE with standard VLE identified additional focal abnormalities in half of the cases with a moderate level of confidence that dysplasia was located. Incorporating IVE into the VLE procedure added only a few minutes of time to each study. These preliminary data suggest that VLE with IVE will further enhance an endoscopist’s ability to identify dysplastic disease without missing a focal lesion, especially for early VLE users.Figure 2View Large Image Figure ViewerDownload Hi-res image Download (PPT)