Genetic algorithm spectral feature selection coupled with quadratic discriminant analysis for ATR-FTIR spectrometric diagnosis of basal cell carcinoma via blood sample analysis

Abstract

A diagnostic approach for basal cell carcinoma (BCC) has been developed based on investigation of the infrared spectra of blood samples. Different predictive procedures were developed using quadratic discriminant analysis (QDA) combined with a simple filtered method and genetic algorithm (GA) as a feature subset and wavelength selection strategy. Results showed 94.74% and 100% accuracy for Filtering-QDA and GA-QDA models, respectively.