Automatic Labeling of Colonoscopy Video for Cancer Detection
The labeling of large quantities of medical video data by clinicians is a tedious and time consuming task. In addition, the labeling process itself is rigid, since it requires the expert’s interaction to classify image contents into a limited number of predetermined categories. This paper describes an architecture to accelerate the labeling step using eye movement tracking data. We report some initial results in training a Support Vector Machine (SVM) to detect cancer polyps in colonoscopy video, and a further analysis of their categories in the feature space using Self Organizing Maps (SOM). Our overall hypothesis is that the clinician’s eye will be drawn to the salient features of the image and that sustained fixations will be associated with those features that are associated with disease states.
KeywordsSupport Vector Machine Cancer Detection Polyp Detection Interest Operator Endoscopy Video
Unable to display preview. Download preview PDF.
- 4.Hwang, S., Oh, J., et al.: Automatic measurement of quality metrics for colonoscopy videos. In: MM’05, Singapore (2005)Google Scholar
- 6.Magoulas, G., Plagianakos, V., et al.: Tumor detection in colonoscopy using the unsupervised k-windows clustering algorithm and neur. networks. In: ESBE (2004)Google Scholar
- 7.Privitera, C., Stark, L.: Algorithms for defining visual regions-of-interest: Comparison with eye fixations. IEEE Trans. on PAMI 22, 970–982 (2000)Google Scholar
- 8.Cheng, W.H., Chu, W.T., Wu, J.-L.: A visual attention based region-of-interest determination framework for video sequences. IEICE Trans Inf. & Syst. E88-D, 1578–1586 (2005)Google Scholar
- 10.Kienzle, W., Wichmann, F., et al.: Learning an interest operator from human eye movements. In: CVPR’06 Workshop (2006)Google Scholar
- 12.Kohonen, T.: Self-Organized Maps. Springer, Heidelberg (1995)Google Scholar
- 13.Jones, M., Viola, P.: Face recognition using boosted local features (2003)Google Scholar