Open AccessA study on the early detection of colon cancer using the methods of wavelet feature extraction and SVM classifications of FTIR
Author(s) -
Cungui Cheng,
Yuan Tian,
Wenying Jin
Publication year - 2008
Publication title -
spectroscopy
Language(s) - English
Resource type - Journals
eISSN - 1875-922X
pISSN - 0712-4813
DOI - 10.1155/2008/182564
Subject(s) - support vector machine , pattern recognition (psychology) , kernel (algebra) , artificial intelligence , feature extraction , feature (linguistics) , fourier transform infrared spectroscopy , radial basis function kernel , wavelet , fourier transform , dysplasia , colorectal cancer , computer science , cancer , mathematics , medicine , kernel method , physics , optics , mathematical analysis , linguistics , philosophy , combinatorics
This paper introduces a new method for the early detection of colon cancer using a combination of feature extraction based on wavelets for Fourier Transform Infrared Spectroscopy (FTIR) and classification using the Support Vector Machine (SVM). The FTIR data collected from 36 normal SD rats, 60 1,2-DMH-induced SD rats, and 44 second generation rats of those induced rats was first preprocessed. Then, 12 feature variants were extracted using continuous wavelet analysis. The extracted feature variants were then inputted into the SVM for classification of normal, dysplasia, early carcinoma, and advanced carcinoma. Among the kernel functions the SVM used, the Poly and RBF kernels had the highest accuracy rates. The accuracy of the Poly kernel in normal, dysplasia, early carcinoma, and advanced carcinoma were 100, 97.5, 95% and 100% respectively. The accuracy of RBF kernel in normal, dysplasia, early carcinoma, and advanced carcinoma was 100, 95, 95% and 100% respectively. The results indicated that this method could effectively and easily diagnose colon cancer in its early stages.