Harnessing the potential of Raman for practical applications

Despite the great potential of Raman spectroscopy for label-free molecular identification and quantification, the widespread use of this optical technique is generally hindered by the fact that it is an exceptionally weak optical process in which only one in a million incident photons experience the Raman scattering.

The scarcity of photons and the necessity to resolve the discrete and narrow Raman lines are the main concerns that need to be properly handled when designing a Raman spectrometer.  Moreover, in order to reduce the strong background of fluorescence while still retaining a reasonable Raman signal, it is customary to implement a so-called confocal optical system together with lens objectives that collect photons over a large solid angle.

Finally, the spectrometer, which is used to disperse the Raman photons by their energy, must feature high throughput and resolving power in order to distinguish the various Raman lines with a good signal-to-noise ratio (SNR). Depending on the required SNR, it might be necessary to cool the detector as to suppress the thermal noise.

Application of Raman spectroscopy in tissue, or in biological samples in general, adds additional challenges to obtaining a useful Raman spectrum. Most importantly, the presence of absorption and scattering of the photons in the tissue lead to reduced Raman signal, distorted spectra, and an increased background fluorescence. Additionally, absorption, which is mainly due to the presence of chromophores and water, limits the choice of laser wavelength to the near-infrared biological window of ~650-1350nm. The laser wavelength is frequently chosen between 700-850nm as a balance between low fluorescence, maximizing the number of generated Raman photons, and allowing the use of silicon-based detectors.

The illustration above visualizes the penetration of near-infrared light in the skin and how photons may undergo scattering events, thus eventually (if not absorbed) leaving the skin with a lateral displacement and at large angles. Furthermore, it is important to note that the dynamic glucose signal resides in the living part of the skin (living epidermis and dermis) and, for this reason, the Raman instrumentation should exclude the signal from the top skin layer (stratum corneum). In practice, this is achieved using the confocal principle that allows for depth selectivity. Last but not least, it should be emphasized that a biological Raman spectrum features a multitude of Raman lines from many different molecules and, consequently, it requires advanced multivariate analysis techniques to quantify the concentration of a specific molecule.

A Raman system is comprised of three central hardware components: an excitation laser, a light guiding probe optics unit and a spectral dispersing element with a detector- a spectrometer. Besides these parts, there are electronics, power electronics, for example, a microcontroller, and a display unit.

The laser is a frequency-stabilized, solid state laser, which is small, inexpensive and rugged. The spectrometer is transmission grating based, analyzing the Raman scattered light from the sample being transmitted via the probe optics.

Every component is optimized and standardized for home use.

In the plot in black, a Raman spectrum of Thenar skin background subtracted and average over 121 spectra with an exposure time of 10 sec is seen. In red a scaled Raman spectrum of glucose/water solution 555.5 mmol/l with an exposure time 13 sec. Both acquired with the WM34 data-mining equipment.

Through our highly skilled team, we have been able to advance the technological application of Raman spectroscopy to be used for non-invasively measuring relevant molecules in human tissue. These advancements have come as a result of nine years of R&D and pre-clinical studies where we refined Raman technology,  focusing on  improving the optical system, electronics, data analysis algorithms, as well as know-how of in-vivo measurements required to:

• Selectively and efficiently collect Raman signal from the interstitial fluid and cells
• Analyze spectral data to derive the concentration of a given molecule
• Build a calibration model and use it for deriving concentration values from measurements

In the process, we have discovered and patented fundamental aspects of Raman detection of glucose and other substances in human tissue. This removes the dependence of signal calibration on probe position, allowing for unprecedented performance.

RSP Systems utilizes a confocal optical design enabling Raman spectroscopy measurements from skin layers below the top layer (Stratum Corneum). By excluding the optical signal from the upper skin layers and lower layers (Reticular Dermis). 

A confocal optical system using a spatial pinhole to block light originating from out-of-focus regions in the sample. The lenses are designed in such a way that the excitation light is focused on the desired depth and light collected from the specific depth is guided through the pinhole. Preferable it is only light originating from the focal plane that is collected.