| abstract |
The present invention relates to a method of training a classification system for characterising a biological sample, a diagnostic classification system, as well as a method of characterising a condition in an animal or a human being by using parameters obtained from the sample. The invention relates to classification based on physical parameters obtained from luminescence spectroscopy on light emitted from the sample. The data obtained from a spectrofluorimetric analysis can be considered a finger-print of the sample. Each sample gives rise to a unique spectrofluorometric set of physical parameters. By analysing the fluorescence data, it is possible to classify samples into two or more classes based on the fluorescence spectra, such as classifying with respect to presence/abs of a specific disease, group of diseases or risk of later attaining a specific disease or a body condition, or concentration of a specific compound or medicine. |