Authors : Mikhaylov Dmitry, Konev Vladimir, Starikovskiy Andrey, Khabibullin Timur and Kuzminova Alla
Abstract: The wireless capsule endoscopy is a method of non-invasive examination of gastrointestinal tract widely used for screening purposes. The review of images taken by capsule usually takes several hours and requires a high degree of concentration by a doctor. To reduce the time and enhance the accuracy of data reviewing the special software for image processing is being developed. It is able to recognize abnormalities of the digestive tract. The study provides the information about the proposed software architecture and its implementation, the function it is to perform, user-friendly report interface development and the proposal for the creation of the digestive diseases’ atlas that will provide information useful for both the professional community of physicians and for the developers of software for detection of pathologies and their symptoms in the images of gastrointestinal tract. The study provides the results of the software testing showing 94% accuracy. The research is underway to increase the accuracy of the algorithm to diversify the pathologies that can be identified in the images obtained by the wireless endoscopic capsule and to implement the atlas of diseases.
Mikhaylov Dmitry, Konev Vladimir, Starikovskiy Andrey, Khabibullin Timur and Kuzminova Alla, 2015. Development of the Software for Recognition of Pathologies in the Images Obtained by the Wireless Endoscopic Capsule and the Atlas of Abnormalities. International Journal of Soft Computing, 10: 279-286.