627 Analysis of High Resolution Microendoscopic Images for the Automated Diagnosis of Colon Neoplasia

This study aimed to develop quantitative HRME image analysis algorithms based on pathological and architectural features. In addition, it aimed to evaluate the automated classification algorithms performance to accurately categorize neoplastic (tubular adenoma, tubulovillous adenoma and cancer) and non-neoplastic (normal, hyperplastic) mucosa in the colon.

Source: Gastrointestinal Endoscopy - April 28, 2015 Category: Gastroenterology Authors: Karen Du, Sharmila Anandasabapathy, Rebecca Richards-Kortum, Yubo Tang, Alexandros D. Polydorides, Shannon Chang Tags: Oral abstract Source Type: research