Method for detecting abnormal cells by using a Fourier transform infrared spectroscopy

摘要

(57) [Abstract] The present invention teaches a method for identifying abnormalities in cells associated with the disease state. In one aspect, the invention is a method for distinguishing malignant and pre-malignant stages of cervical cancer from normal cervical cells. The method utilizes infrared cervical cells exfoliated dried infrared permeable matrix on scanning in the frequency range of the 3000~950cm ^-1 (IR) spectrum. Identification of the sample is based on establishing a tested using a representative set of spectra of malignant sample normal, dysplastic and. During the course of the test, multivariate techniques (PCA) and / or partial least squares, such as (PLS) principal component analysis is used. PLS and PCA can form clusters in a multidimensional space by reducing the data based on the maximum change between spectra, showing the different populations. (Principal component regression based on the data that has been reduced from, for example PCA) use linear regression to form the basis for identification or Mahalinobis distance. It is used easily, this method, cervical smear of the following groups: do (individuals with no prior history of dysplasia), a distinction that can be statistically reliable sample between malignant and normal dysplasia. The present invention discloses a method of using a spectrum of individual cells in the above method and the method of obtaining the IR spectrum of the individual cervical cell fixed infrared transparent matrix. In a further aspect, the present invention is a method for using vibrational spectroscopy imaging to distinguish between cells of normal and disease.