Assisted differentiated stem cell classification in infrared spectroscopy using auditory feedback

In this study we investigate ways in which data sonification can improve standard data analysis techniques currently employed in the analysis of stem-cells using Fourier Transform Infrared (FTIR) Spectroscopy. Four different sonification methods have been evaluated and their effectiveness has been evaluated through listening tests, designed to assess the discriminating capability of the auditory technique. We identify FM synthesis driven by feature extraction as the most perceptually relevant technique for the auditory classification of FTIR data. Whilst this technique is not commonly used in sonification research, it allows us to utilise the most salient characteristics of the absorption spectra, leading to an improved classification accuracy with a clear timbral differences between differentiated and non-differentiated cell-types.