Gastroesophageal Reflux Disease Diagnosis Using Hierarchical Heterogeneous Descriptor Fusion Support Vector Machine

A new computer-aided diagnosis method is proposed to diagnose the gastroesophageal reflux disease (GERD) from endoscopic images of the esophageal-gastric junction. To avoid the interferences of different endoscope devices and automatic camera white balance adjustment, heterogeneous descriptors compu...

Full description

Saved in:
Bibliographic Details
Published in:IEEE transactions on biomedical engineering 2016-03, Vol.63 (3), p.588-599
Main Authors: Huang, Chun-Rong, Chen, Yan-Ting, Chen, Wei-Ying, Cheng, Hsiu-Chi, Sheu, Bor-Shyang
Format: Article
Language:English
Subjects:
Computational modeling
Computer aided diagnosis
Diseases
Endoscopes
Endoscopy
Endoscopy, Gastrointestinal - methods
Feature extraction
Gastroesophageal reflux
Gastroesophageal Reflux - diagnosis
Humans
Image color analysis
Image Interpretation, Computer-Assisted - methods
Junctions
Lighting
Support Vector Machine
Support vector machines
Citations: Items that this one cites
Items that cite this one
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:A new computer-aided diagnosis method is proposed to diagnose the gastroesophageal reflux disease (GERD) from endoscopic images of the esophageal-gastric junction. To avoid the interferences of different endoscope devices and automatic camera white balance adjustment, heterogeneous descriptors computed from heterogeneous color models are used to represent endoscopic images. Instead of concatenating these descriptors to a super vector, a hierarchical heterogeneous descriptor fusion support vector machine (HHDF-SVM) framework is proposed to simultaneously apply heterogeneous descriptors for GERD diagnosis and overcome the curse of dimensionality problem. During validation, heterogeneous descriptors are extracted from test endoscopic images at first. The classification result is obtained by using HHDF-SVM with heterogeneous descriptors. As shown in the experiments, our method can automatically diagnose GERD without any manual selection of region of interest and achieve better accuracy compared to states-of-the-art methods.
ISSN:0018-9294
1558-2531
DOI:10.1109/TBME.2015.2466460