Hyperspectral imaging in biophysics and energy physics

Over the past 30 years, spectroscopic techniques, such as IR, near IR, and Raman spectroscopy have become very popular. They allow investigations of composed and heterogeneous materials in their native forms without destroying the biological matrices of cells and tissues or the solid-state properties of materials. Prevailing research within this field is the development of spectral pre-processing techniques, especially the estimation of so-called Mie scattering in single cell infrared microspectroscopy and the development of radiative transfer models for coupled, turbid media. Scattering effects have been considered as a major obstacle for the interpretation and further use of the infrared spectra and hyperspectral images. The project group has developed a theory that describes analytically Mie scattering relevant for infrared microscopes with high numerical apertures as used in the infrared microspectroscopy of cells and tissues. The obtained analytical understanding of the problem has been used to develop a method for correcting Mie type scattering in infrared spectroscopy of single cells. The correction of Mie type scattering in spectra of single cells is an important prerequisite for the interpretation of the chemical absorbance bands in the spectra. The developed method has attracted attention in the community of biomedical infrared spectroscopy. A scattering phenomenon that appears frequently in infrared spectroscopy and hyperspectral imaging of tissues are the so-called fringes. Fringes are interference effects that appear due to internal reflections in the material. They are present in many applications of infrared spectroscopy and hyperspectral imaging such as infrared microspectroscopy of biological tissues or life cell imaging. During the project, the group developed a method for the correction of fringes in tissue spectra and infrared images. The investigation of inherent optical properties of turbid or random media is an important requisite for the estimation of the amount and type of algae species in oceans. In order to develop a model describing the inherent optical properties of turbid or random media, it was investigated how a representation of the inherent optical properties of non-spherical particles can be obtained by a collection of spherical particles, where each sphere has the same volume-to-surface ratio as the non-spherical particle. In particular, scattering by spherical and non-spherical particles using Mie theory codes was considered. By this approach, it was possible to analyse scattering by spherical particles and to analyse scattering by randomly oriented non-spherical particles by T-matrix codes. Fifteen algae species representative of Norwegian coastal water were investigated using the developed code. The project had a close collaboration with several international groups: with Prof. Dr. Reinhold Blümel from the Department of Physics, Wesleyan University, Middletown, CT, US on Mie scattering, with Prof. Hirshmugl at the University of Milwaukee, Wisconsin on hyperspectral imaging, and with the research group of Prof. K. Stamnes at Stevens Institute of Technology, USA on radiative transfer models. The project led to several spin-off projects: The PostDoc in Ås won a Unge Forskertalenter project entitled -No: 250678 /F20. The project group was invited to participate in the NSF Grant managed by Prof. Hirshmugl at the University of Milwaukee, Wisconsin which follows up the activity on infrared spectroscopy/hyperspectral imaging of cells and tissues. The results related to the understanding of light scattering in the infrared microspectroscopy/hyperspectral imaging have a strong impact in the field of biomedical vibrational spectroscopy, where our methods are used for data processing on a routine basis. The newly developed activity related to light management in solar cells with nanostructures has high importance for the master program in energy physics at IMT/NMBU. The results related to inherent optical properties of non-spherical particles are of great interest in ocean optics for modelling scattering and absorption by algae and in medical optics for modelling scattering and absorption by tissue. The results related to radiative transfer modelling with a rough interface, including polarization, are of great interest in optical remote sensing of the atmosphere-ocean system, and in medical optics for optical remote sensing of tissue. Results were disseminated by scientific publications in prestigious journals and on conferences within physics, biology and biomedical communities.

With this proposal, the Department of Mathematical Sciences and Technology (IMT) at the Norwegian University of Life Sciences (UMB) follows up the recent evaluation of physics research in Norway, which recommended focusing on and strengthening biophysics and energy physics research at UMB. We propose a common research platform integrating the biophysics and energy physics groups at IMT to develop and apply innovative methods in spectroscopy and hyperspectral imaging. The Optics and Atomic Physics group at the Department of Physics and Technology, University of Bergen (UiB), is a partner in this project. Spectroscopy and hyperspectral imaging are powerful tools for the non-destructive investigation of living and non-living materials, providing spectral a nd spatial information of the sample at macroscopic and microscopic scales. At UMB, we are currently using these tools to investigate multicrystalline silicon wafers for solar cell applications and to develop pioneering applications of spectroscopy of str ongly scattering biological samples such as cells and tissues for cancer diagnostics and treatment. At UiB, these tools are used to develop novel radiative transfer models of coupled, turbid media, with potential applications in cancer detection. Comprehe nsive expertise in multivariate data analysis at UMB will make a crucial contribution to the analysis of the enormous amounts of data generated by hyperspectral imaging techniques. Progress in biophysics and energy physics will benefit greatly from a jo int platform in spectroscopy and hyperspectral imaging. In particular, this platform will focus on optimizing data acquisition, understanding the nature and limitations of the data and developing strategies for processing and interpretation. We will thu s develop a national competence centre for cutting-edge experimental and analytical techniques applicable to a wide range of research fields and applications, from solid-state physics to cancer diagnostics.