HUMAN GASTROINTESTINAL DATA ANALYSIS BASED ON WAVELET TRANSFORM-FRACTAL ALGORITHM CHARACTERIZED BY WIRELESS CAPSULE

摘要

Background: Gastrointestinal tract pressure information can reflect the dynamic disorder of gastrointestinal tract physiological and pathological changes, which is important in early diagnosis and treatment of gastrointestinal diseases. However, the rule of human gastrointestinal activities under his normal state is still not clear. The purpose of this paper is to find some rules of gastrointestinal activities using wavelet transform-fractal algorithm and provides a reference for the diagnosis of gastrointestinal disease. Methods: Data from 16 healthy subjects (13 men and three women) aged 18-35 years were analyzed. The key challenges on the data processing were filtering, features extraction, information evaluation and providing the reference conclusion which were resolved by employing wavelet decomposition, fractal algorithm etc. Key Results: About 94% of the subjects gastrointestinal activities follow the rule that the colonic activities in the digestive period are the roughest with the largest box dimension, closely followed by small intestinal activities with gastric activities ranking third with the smallest box dimension, which is never revealed before. Moreover, for the same person the gastrointestinal activities in the digestive period are rougher than in the inter- digestive period, as the box dimension of gastrointestinal activities in the digestive period is larger than in the inter- digestive period. Conclusions and Inferences: The results show that the wavelet transform-fractal algorithm provides a new way to research human gastrointestinal activities.