After a general overview on the various types of features extracted and on the main selection and classification methods implemented in the frame of tissue fibered spectroscopy studies, the present contribution will focus on the implementation of multi-class classification schemes and the original combination of decision fusion approaches based on AutoFluorescence (AF) and Diffuse Reflectance (DR) modalities and spatial resolution through multiple Collecting-Excitation Fiber Distances (CEFS). Results confirm the interest to couple AF and DR spectroscopy measurements and show that this innovative solution based on spatial resolution data fusion leads to optimized discrimination between histological classes and reduced confusion between the closest histological classes.