Related Experiment Video
Updated: Dec 28, 2025
15:17
Colorectal Cancer Cell Surface Protein Profiling Using an Antibody Microarray and Fluorescence Multiplexing
Published on: September 25, 2011
14.3K
Polyp fingerprint: automatic recognition of colorectal polyps' unique features.
Ana García-Rodríguez1, Jorge Bernal2, F Javier Sánchez2
1Endoscopy Unit, Gastroenterology Department, Hospital Clínic, IDIBAPS, CIBEREHD, University of Barcelona, Barcelona, Spain.
Surgical Endoscopy
|February 13, 2020
Summary
A new content-based image retrieval (CBIR) system accurately identifies the same colon polyp in 91% of cases using machine learning. This polyp fingerprinting aids in accurate polyp image recognition during colonoscopies.
Area of Science:
- Medical Imaging
- Machine Learning
- Gastroenterology
Background:
- Content-based image retrieval (CBIR) utilizes machine learning to find similar images based on features.
- Developing a CBIR system to identify unique polyp features ('polyp fingerprint') is crucial for accurate diagnosis.
Purpose of the Study:
- To develop and evaluate a CBIR system for identifying images of the same colon polyp.
Main Methods:
- Employed the Bag of Words machine learning technique to uniquely describe each endoscopic polyp image.
- Tested the system on 243 white light images from 99 distinct polyps captured during routine colonoscopies.
Main Results:
- The CBIR system achieved a 91% accuracy in matching images of the same polyp.
- Performance was consistent across different polyp classifications (Paris classification) and sizes (<10 mm vs. >10 mm).
Conclusions:
- A CBIR system can accurately match images of the same polyp, serving as a valuable tool for polyp recognition in clinical settings.
