Prospective Evaluation of Multi-modal Optical Imaging with Automated Image Analysis to Detect Oral Neoplasia In Vivo

The five-year survival rate for patients with oral cancer remains low, in part because diagnosis often occurs at a late stage. Early and accurate identification of oral high-grade dysplasia and cancer can help improve patient outcomes. Multi-modal optical imaging is an adjunctive diagnostic technique in which autofluorescence imaging is used to identify high-risk regions within the oral cavity, followed by high resolution microendoscopy to confirm or rule out the presence of neoplasia. Multi-modal optical images were obtained from 206 sites in 100 patients. Histologic diagnosis, either from a punch biopsy or an excised surgical specimen, was used as the gold standard for all sites. Histopathologic diagnoses of moderate dysplasia or worse were considered neoplastic. Images from 92 sites in the first 30 patients were used as a training set to develop automated image analysis methods for identification of neoplasia. Diagnostic performance was evaluated prospectively using images from 114 sites in the remaining 70 patients as a test set. In the training set, multi-modal optical imaging with automated image analysis correctly classified 95% of non-neoplastic sites and 94% of neoplastic sites. Among the 56 sites in the test set that were biopsied, multi-modal optical imaging correctly classified 100% of non-neoplastic sites and 85% of neoplastic sites. Among the 58 sites in the test set that corresponded to a surgical specimen, multi-modal imaging correctly classified 100% of non-neoplastic sites and 61% of neoplastic sites. These findings support the potential of multi-modal optical imaging to aid in the early detection of oral cancer.

from # All Medicine by Alexandros G. Sfakianakis via Alexandros G.Sfakianakis on Inoreader http://ift.tt/2vgW85t
via IFTTT