Raman spectroscopy (RS) is a noninvasive label-free spectroscopic technique that probes molecular information of a sample with light (Ferraro and Nakamoto, 2012). The contrast of RS comes from the inelastic scattering (Raman scattering) of photons by the molecules, where a gain or loss of energy occurs, leading to a frequency shift of the scattered light. In RS, the Raman scattering is usually analyzed with respect to wavenumber, where the peaks of the spectrum serve as fingerprint identifications of different molecules (Ferraro and Nakamoto, 2012). The time required to produce the Raman spectrum is generally within seconds to minutes, allowing for a rapid biochemical analysis. Near-infrared light ranging from ~700 nm to ~1064 nm has been most commonly used to illuminate the biological samples in RS due to the auto-fluorescence in the visible light range and the photo-degradation of tissue by the light in the ultra-violet spectral range (Movasaghi et al., 2007).
Imaging with RS can provide high-resolution 3-D distribution of the specific molecules in a heterogeneous biological sample by spatially mapping the Raman spectrum information (Zhang et al., 2010b). The spatial scale of RS imaging of cells and tissues is similar to the traditional optical microscopy but with reduced resolution and higher penetration depth due to the larger wavelength of light that is employed. Specifically, the spatial resolution can reach the submicron level and is limited by optical diffraction (Zhang et al., 2010b). Utilizing a confocal configuration, confocal RS imaging has an improved spatial resolving capability and is able to provide depthwise sectioning with a penetration depth of up to the millimeter level (Dieing et al., 2011). Hyperspectral RS imaging data can be achieved with both point scanning and line scanning schemes for 3-D molecular imaging with high specificity (Zoubir, 2012).
Since Raman scattering is significantly weaker compared with Rayleigh scattering and fluorescence, traditional RS might suffer from low signal-to-noise ratio, thus requiring a longer time of data collection. The development of Raman scattering-based spectroscopic methods has led to a number of advanced RS techniques, such as surface-enhanced resonance RS (McNay et al., 2011) and coherent anti-Stokes RS (Evans and Xie, 2008), which provide increased Raman scattering and improved resolution and sensitivity. Although molecular specific imaging can be achieved with fluorescence labeling in CM and MPM, the unique features of RS that include utilizing endogenous contrast, requiring minimum sample preparation, and simultaneously obtaining a large amount of molecular composition data are very attractive and make RS a useful molecular imaging tool for both cells and tissue constructs in tissue engineering (Jell et al., 2010; Perlaki et al., 2014).