An Application of Multivariate Data Analysis to Photoacoustic Imaging for the Spectral Unmixing of Gold Nanorods in Biological Tissues

Abstract

Gold nanorods (GNRs) showed to be a suitable contrast agent in photoacoustics (PA), and are able to provide a tunable absorption contrast against background tissue, while a detectable PA signal can be generated from highly localized and targeted areas. A crucial issue for these imaging techniques is represented by the discrimination between exogenous and endogenous contrast and the assessment of the real PA signal magnitude. The application of image resolution/unmixing methods was implemented and optimized to recover the relative magnitude spectra and distribution maps of image constituents of the biological sample based on multivariate analysis (multivariate curve resolution—alternating least squares, MCR-ALS) in the presence of GNRs with tunable absorption properties. The proposed data analysis methodology is demonstrated on real PA images from experimental animal models and ex-vivo preparations.