Discrete Wavelet Transform based colon polyp detection using synthesize similarity measure
Mainak Biswas, Sharmadeb Jana, Sneha Bhattacharya, Sayid Aktar, Debasis Maji, Sovan Ghosh, Swarnendu Patra, Gautam Sarkar · 2017 1st International Conference on Electronics, Materials Engineering and Nano-Technology (IEMENTech) · 2017
The most the most dangerous cancer is colon cancer. The premature version of this type of cancer is the colon polyp. The death ration for this type of cancer is almost 10:9.2 for men: women [1]. Only one prevention of this is the very early diagnosis. In this paper a robust and novel work is proposed for detecting this colon cancer using Discrete Wavelet Transform (DWT). For this detection two sets of videos are taken from American College of Gastroenterology [3], actually these two videos are fissure into number of image samples. A novel Wavelet Transform is applied on each image sample followed by edge detection technique using Otsu edge detection process. For extracting the features of those images Principle Component Analysis (PCA) is executed for reduction of the features. A potent classification, Synthesize Similarity Measure (SSM) algorithm is opted for the classifying the disease. This fast and ameliorated process (DWT_SSM) gives the best result (almost 99.1% best result, 98.82% average result) when comparing with other benchmark and conventional procedures.