Non-negative Matrix Factorization for Endoscopic Video Summarization

Authors:
Spyros Tsevas;Dimitris Iakovidis;Dimitris Maroulis;Emmanuel Pavlakis;Andreas Polydorou
Affiliations:
Dept. of Informatics and Telecommunications, University of Athens, Panepistimiopolis, Athens, Greece GR-15784;Dept. of Informatics and Telecommunications, University of Athens, Panepistimiopolis, Athens, Greece GR-15784;Dept. of Informatics and Telecommunications, University of Athens, Panepistimiopolis, Athens, Greece GR-15784;Department of Surgery, Aretaieion Hospital, Athens, Greece 115 27;Department of Surgery, Aretaieion Hospital, Athens, Greece 115 27
Venue:
SETN '08 Proceedings of the 5th Hellenic conference on Artificial Intelligence: Theories, Models and Applications
Year:
2008

Citing 9
Cited 0

Detecting Abnormal Patterns in WCE Images

BIBE '05 Proceedings of the Fifth IEEE Symposium on Bioinformatics and Bioengineering
ROC curves and video analysis optimization in intestinal capsule endoscopy

Pattern Recognition Letters - Special issue: ROC analysis in pattern recognition
Automatic Detection of Intestinal Juices in Wireless Capsule Video Endoscopy

ICPR '06 Proceedings of the 18th International Conference on Pattern Recognition - Volume 04
A neuro-fuzzy-based system for detecting abnormal patterns in wireless-capsule endoscopic images

Neurocomputing
Fast communication: Unsupervised data reduction

Signal Processing
Automatic classification of digestive organs in wireless capsule endoscopy videos

Proceedings of the 2007 ACM symposium on Applied computing
A machine learning framework using SOMs: applications in the intestinal motility assessment

CIARP'06 Proceedings of the 11th Iberoamerican conference on Progress in Pattern Recognition, Image Analysis and Applications
Nonnegative Lagrangian relaxation of k-means and spectral clustering

ECML'05 Proceedings of the 16th European conference on Machine Learning
MPEG-7 Visual Descriptors—Contributions for Automated Feature Extraction in Capsule Endoscopy

IEEE Transactions on Circuits and Systems for Video Technology

Quantified Score

Hi-index	0.00

Visualization

Abstract

Wireless Capsule Endoscopy (WCE) has been introduced as a non-invasive colour imaging technique for the inspection of the small intestin along with the rest of the gastrointestinal tract. Each WCE examination results in a 50,000-frames video that has to be visually inspected frame-by-frame by the doctor and this may be a highly time-consuming task even for the experienced gastroenterologist. In this paper we propose a novel approach that leads to a summarized version of the original video enabling significant reduction in the video assessment time without losing any critical information. It is based on symmetric non-negative matrix factorisation initialized by the fuzzy c-means algorithm and it is supported by non-negative Lagrangian relaxation to extract a subset of video frames containing the most representative scenes from an entire examination. The experimental evaluation of the proposed approach was performed using previously annotated endoscopic videos from various sites of the small intestine.