Blind Source Separation of Photoacoustic Depth Profiles into Independent Components

Step-scan photoacoustic spectroscopy is a powerful tool to nondestructively retrieve depth related information from a sample. Through digital signal processing a series of spectra with effectively different modulation frequencies, probing different thermal diffusion lengths within a sample, can be collected simultaneously. For layered samples spectra of the constituent layers can then be obtained by calculating spectra at specific phase angles from the in-phase and quadrature data through phase projection. However, without prior knowledge of the spectra of the constituent layers, this approach can be difficult. In this report we present an alternate possibility for evaluating step scan photoacoustic data, namely independent component analysis (ICA), which allows for “blind separation” of the mixed photoacoustic spectra without prior knowledge of the constituent spectra. Phase projection and ICA are applied to photoacoustic data acquired from a multilayer sample in an attempt to isolate the spectra of the constituent layers. The results for the two methods are comparable, with ICA offering the advantage that no prior information about the pure spectra of the sample layers is needed.