Improving the Quality of Color Colonoscopy Videos
- PDF / 1,717,089 Bytes
- 8 Pages / 600.05 x 792 pts Page_size
- 92 Downloads / 253 Views
Research Article Improving the Quality of Color Colonoscopy Videos ˜ and Gerard Lacey Rozenn Dahyot, Fernando Vilarino, Department of Computer Science, School of Computer Science and Statistics, Trinity College Dublin, College Green, Dublin 2, Ireland Correspondence should be addressed to Rozenn Dahyot, [email protected] Received 1 August 2007; Revised 20 November 2007; Accepted 22 January 2008 Recommended by Shoji Tominaga Colonoscopy is currently one of the best methods to detect colorectal cancer. Nowadays, one of the widely used colonoscopes has a monochrome chipset recording successively at 60 Hz R, G, and B components merged into one color video stream. Misalignments of the channels occur each time the camera moves, and this artefact impedes both online visual inspection by doctors and offline computer analysis of the image data. We propose to restore this artefact by first equalizing the color channels and then performing a robust camera motion estimation and compensation. Copyright © 2008 Rozenn Dahyot et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
1.
INTRODUCTION
Colorectal cancer is the second leading cause of cancer death in the United States and colonoscopy, by removing polyps early, is currently one of the best methods to reduce this fatality [1]. Colonoscopy is a minimally invasive endoscopic examination of the colon and the distal part of the small bowel with a fiber optic camera on a flexible tube. The video is inspected in realtime by the doctors to give a visual diagnosis (e.g., ulceration, polyps). This procedure also gives the opportunity for biopsy of suspected lesions. The quality of endoscopic screening is of significant concern in the medical community. Large interendoscopist variation in the number of polyps being missed has been measured in clinical studies [1]. Although no definitive cause for the high miss rates has been identified, the speed of camera movement has been suggested as a cause. Our research is within this context of identifying image quality artefacts that may be contributory factors to the high incidence of miss rates in endoscopy. The inspection of colonoscopy videos can also be done offline, and computer aided methods are currently developed to assist medical doctors. For instance, in [2], a method is proposed to detect tumors in colonoscopy videos using color wavelet covariance and linear discriminant analysis. In [3], the video is used to assess the endoscopist’s skills by esti-
mating the camera motion. In [4], edge detection and region growing are used to help the control of the colonoscope. In [5], an automatic labeling system for colonoscopy videos is presented using eye tracking of experts for training and indexing purposes. Labeled data is then used to feed a support vector machine classifier to automatically detect tumors. Endoscopes used in hospital use different imaging systems. Inde
Data Loading...
