Artificial intelligence-based analysis of surface enhanced Raman scattering spectrum of Candida albicans

Objective To construct an AI data preprocessing framework based on the Raman strength data of Candida albicans (C. albicans), in order to facilitate subsequent prediction using convolutional neural networks (CNN) models.Methods C. albicans were isolated from infected patients, and positively charged silver nanopartides were used as base fluid for collection of surface enhanced Raman scattering (SERS) spectrum. First, extremum in the collected data were eliminated, and then Fourier transform was used to convert data from the time domain into the frequency domain to remove environmental noise. After the smoothed data was normalized, it was converted into the value of polar coordinates, and the value of polar coordinates was upgraded to the dimension of the original data by the Gramey angle field.Results After the processed C. albicans were converted into heat maps, the data dimension was improved to facilitate the classification of fungal types.Conclusions The data preprocessing framework of Raman intensity of C. albicans is successfully established, which provides a very reliable tool for the prediction of SERS spectrum of C. albicans.