Nowadays, the diagnosis of small bowel motility dysfunctions is basically performed by using minimally invasive techniques, which can only be conducted at some referral centers around the world. Wireless Video Capsule Endoscopy (WVCE) represents a novel diagnostic tool that presents clinical advantages, since it is non-invasive and at the same time it provides, for the first time, a full picture of the small bowel morphology, contents and dynamics.

The presence or absence of diverse physiological symptoms constitutes the first evidence for the diagnosis of a given pathology of the small bowel. However, nowadays, the main source of information, and the only one which leads to a conclusive positive diagnosis of intestinal motility disorders, is that obtained from the result of the motility test performed by using manometric devices. This method consists of recording pressure values along a set of points in the proximal small intestine. However, this technique shows several drawbacks which constitute a problem for its generalized use. Manometry is a complex technical procedure that must be restricted to those centers where the specialized staff and technology are present. In addition, manometry is considered a minimally invasive diagnostic method, but it involves the introduction of a 1.5m tube through the patient’s esophagus with the associated discomfort for the patient and the need for his/her hospitalization. Finally, motility information provided by manometry is restricted to pressure values exclusively, lacking of information about different content, structure, morphology and dynamics of the intestine.

In this project, we are developing a novel methodology which can potentially represent a shift in the clinical procedures used so far for the assessment of small bowel dysfunctions. We propose a computer-aided system for the analysis of the images provided by wireless capsule endoscopy videos in order to obtain a diagnosis in terms of presence of intestinal dysfunctions. Our proposal is based on three pillars that provide a completely new methodology, namely: 1) The use of data obtained from wireless capsule endoscopy videos, 2) The analysis of a rich set of physiological phenomena that could be related to intestinal motility by means of the use of features calculated through classical Computer Vision techniques, and, finally, 3) The use of statistical machine learning methods for the categorization and analysis of clinical studies in terms of positive and  negative evidences for motility dysfunctions.

Acknowledgments

This project is developed in collaboration with the Hospital Universitari de la Vall d’Hebron, Universitat Autònoma de Barcelona, and is supported by Given Imaging Ltd. (Israel).

Publications

Vilarino, F.; Spyridonos, P.; DeIorio, F.; Vitrià, J.; Azpiroz, F.; Radeva, P.;Intestinal Motility Assessment With Video Capsule Endoscopy: Automatic Annotation of Phasic Intestinal Contractions, Medical Imaging, IEEE Transactions on , vol.29, no.2, pp.246-259, Feb. 2010.

M.  Drozdzal, L.  Igual, P. Radeva, J. Vitrià, C. Malagelada and F. Azpiroz. Aligning Endoluminal Scene Sequences in Wireless Capsule Endoscopy. IEEE Computer Society Workshop on Mathematical Methods in Biomedical Image Analysis MMBIA10. 2010 IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops (CVPRW).