Caveolae image analysis for pathogen diabetes

This paper proposes a hybrid model to detect the caveolae from the high noise mesangial cell image, which can be used to analyze the pathogen diabetes further. The model combines the automated seeded region growing, self-adaptive canny algorithm, morphological techniques. The experiments show that caveolae can be segmented out, and be represented as the suitable descriptors used for further data mining steps. Thus, we can determine the relationships between caveolin description and diabetes through image mining.