Contribution of the LUT University to the infrastructure includes development and establishment of an open-access quantitative spectral data processing and image generation platform. The user-friendly data-analysis tools will be installed on a freely accessible online platform webpage built during the project.
The platform is essential for optimal research outputs with the imaging equipment and is not known to exist. It will address four key requirements: a) big data handling, b) quantitative coherent Raman and IR spectral data analysis, c) user-friendly and flexible image generation, and d) remote access. The data analysis algorithms will be tailored to coherent Raman (CARS and SRS) and near-field IR spectral data. Crucially, quantitative imaging with CARS data has been largely prevented by the nonlinear and coherent nature of the CARS signal. Prof. Vartiainen’s group at LUT have overcome this barrier by pioneering specific wavelet and numerical phase retrieval approaches to extract the linear vibrational contribution in the multiplex CARS signal (Raman line-shape). These algorithms will be incorporated into the analytical platform. Quantitative analysis of these processed CARS, SRS and near-field IR data will be optimized and developed with fully automatic and unsupervised algorithms (including e.g. multivariate curve resolution (MCR)), such that they can be readily applied and used to generate chemical and structural images by end-users. Furthermore, processing of correlative multimodal SFG, TPEF and conventional confocal fluorescence signals will also be optimized.