A wavelet-based approach to detect bladder tumor in color images

A wavelet-based method is proposed to extract the tumor region from a color medical image. The diagnosis of tumors is usually treated by inspecting the color of a tissue section stained by methylene blue. Traditional treatment of this process relied primarily on manual operation, erroneous assessment may thus arise from subjective judgement and eye measurement. Instead of a manual process, we propose a radial search technique, which will improve the accuracy by extracting the border of the tumor. Experimental results show that the proposed method is fast and can extract the tumor region successfully. Moreover, the technology developed in of this research will be helpful in the development of a computer-aided image analysis system to provide an automatic visual feature extraction tool for tumor diagnosis