Multispectral Fluorescence Imaging for Tumor Detection and Molecular Biology

摘要

Optical techniques, such as fluorescence imaging, are of particular interest for visualization of various superficially located epithelial tissues, such as the skin or the mucosa of interior hollow organs including easily reachable areas, such as the oral cavity or genital tract, besides the endoscopically accessible organs. The feature that fluorescence changes early in the development of certain types of malignant lesions and the utility for identification of premalignant lesions is of particular clinical interest. Obvious advantages with fluorescence detection are the minimally invasiveness and real-time aspect. In the clinic this means that LIF can be utilized interactively during the procedure and give up-dated information during the diagnostic procedure. Spectroscopic characterization is based on very early biochemical as well as morphological changes in the tissue. Multispectral fluorescence imaging can also be used for identification of early lesion and for delineating tumours based on exogenous fluorescent tumour markers. For tumor detection located inside solid organs, such as liver, kidney and breast parenchyma, fluorescence techniques are not equally favorable and straight forward due to poor penetration and multiple scattering of light in tissue. A currently very interesting field of research for deep lesion characterization is based on fluorescence mediated tomography, of interest especially for longitudinal small animal investigations, where a fluorescence labeled lesion can be studied for treatment evaluation; as well as for studies of molecular interactions in vivo. Also fluorescence-based optical mammography is presently investigated by several groups. As the tomographic reconstruction is complex and relatively mathematically ill-conditioned, any additional knowledge is welcome to improve the reconstruction and can potentially speed it up and/or make it more robust. The data presented here suggest that multispectral detection can be used for both these purposes.