Estudo comparativo de técnicas de processamento de imagens em imagens da mucosa intestinal

Abstract

Diabetes is a disease that has been much discussed in recent times due to its rapid growth and consequences. It has many classifications, i.e., diabetes mellitus type 1 (DM1), type 2 (DM2), gestational diabetes and other specific cases. The DM2 is the most common type to happen and researches discovered that there is a relationship between changes in the intestinal mucosa and development of this disease. This variation can be detected by medical exams, e.g., digestive endoscopy. As this analysis is done in a visual way, this work proposes a study of techniques of digital image processing (DIP) to prepare the image for further feature extraction and knowledge discovery. For development work, it has been divided into stages, each of which corresponds to the application of a technique DIP. At the end, the results from the experiments made show that was not possible to identify all intestinal villus through of the techniques used.