Team 26

Software System for Measurement of Colon Polyps from Endoscope Feed

Team Members: Teleah Hancer
Mario Mosqueda Crespo
Alexandra Richey
Georgia Sullivan

Team Mentors: Vikram Kodibagkar, PhD - SBHSE
Jewel Samadder, MD - Mayo Clinic

YouTube Link: View the video link below before joining the zoom meeting

Zoom Link: https://asu.zoom.us/j/84557399303

Abstract

Every year, ~19 million colonoscopies are performed in the US. These are often performed to screen for cancerous polyps. Polyp diameter is currently the best metric for detecting cancer and is an important aspect of treatment decisions, but physicians struggle to accurately estimate it using visual estimation. One study showed that physicians can only estimate polyp diameter to within 1mm 46% of the time. Our team's goal is to create a system that can accurately measure the size of polyps during a colonoscopy and aid physicians. In order to address this need, we consulted with Dr. Jewel Samadder, a gastroenterologist at Mayo Clinic, and Dr. Vikram Kodibagkar, an image analysis expert. We have prototyped an image analysis software system that will integrate well with current colonoscopy procedures. The system characterizes the size of a polyp selected by the user from a still frame of the endoscope video feed. The image is calibrated from a reference measurement taken from the forceps used in colonoscopies. Our system was able to achieve an accuracy of over 90% in measuring the longest axis of the polyp and an accuracy of over 75% in measuring the cross-sectional area of the polyp using images from a model colon. Our software is already more accurate than visual estimation alone in estimating diameter, and we are working to increase the accuracy in further iterations of the software. In the future, we hope to be able to test our software on human patients and seek FDA approval.