Fuzzy Clustering for the Classification of Cancerous Cells using FTIR Spectroscopy

Cancer has become a major adversary to human health and the development and enhancement of techniques for use in its diagnosis and treatment have quickly become a focus of worldwide research. In recent years, Fourier transform infrared (FTIR) spectroscopy has been increasingly applied to the study of biomedical conditions and is becoming a powerful tool for determining the biochemical composition within a biological system. In order to analyse the FTIR spectroscopic data from tissue samples, multivariate clustering techniques have often been used to separate sets of unlabelled infrared spectral data into different clusters based on their characteristics. The purpose of clustering is to group the spectral data such that the data in the same cluster are as similar as possible and data within different clusters are as dissimilar as possible. Hence, different types of cells can be separated within biological tissue. Among existing clustering techniques, it has been shown that fuzzy clustering techniques such as fuzzy c-means can have clear advantages over crisp and probabilistic clustering methods, and have been widely used in medical diagnosis and pattern recognition. In this Chapter, we summarise the fuzzy clustering techniques which we have developed and successfully applied to the identification of cancer cells using FTIR spectroscopy in a selection of tissue samples which have kindly been provided by Derby General Hospital and Gloucestershire Royal Hospital, England, UK.