A technique based on wavelet and morphology transform to recognize the cancer cell in pleural effusion

This paper analyzes the cause of cells fallen off into pleural effusion, and its effect on diagnosis of lung cancer. According to features of cancer cells in morphology and structure, the responding presentations in wavelet analysis and morphology are discussed. Some gray-scale features and gray-scale gradient features based on wavelet analysis, and some morphology features about edge intensity are presented. Based on these features a backpropagation neural network is constructed to recognize cancer cells fallen off into pleural effusion. Experimental results show that this method has a high recognition ratio.