Brillouin imaging has recently emerged as a powerful technique for its ability to give insight to the mechanical properties of biomaterial. It exploits inelastic scattering of light by acoustic vibrations and maps the tissue stiffness point by point with micron resolution. The non-invasive, real-time nature of the measurements also makes it a potent candidate for in-vivo imaging of live cells and tissues. This, however, has to rely on a compact and flexible apparatus, a Brillouin endoscope, for remote access to specimen parts.
One of the main challenges encountered in the construction of Brillouin endoscope is that the inelastic scattering in the fibre conduit itself is orders of magnitude stronger than the Brillouin signal scattered by the specimen. This is because the length of the fibre endoscope (meters) is orders of magnitude larger than the imaging volume (microns). The problem can be overcome if the scattered light is collected by a separate fibre and does not mix with the fibre scattering inside the delivery channel.
Here we present an all-fibre integrated Brillouin microspectroscopy system that exploits the paths separation between delivery and collection channels. The experimental setup consists of a pair of standard silica single-mode fibres coupled to a graded-index lens and illuminated with a 671nm continuum wavelength source. We test our system performance on liquid samples of water and ethanol and confirm Brillouin shifts of 5.9 GHz and 4.6 GHz, respectively. More importantly, we do not observe any signals corresponding to Brillouin shift in the fibre, in agreement with expectation.
Mechanical characterisation of biomaterials provides the basis for investigating disease-related changes in the biomechanical properties of living tissues and cells. Brillouin microscopy offers a non-invasive and label-free method to measure material properties. Briefly, Brillouin scattering involves energy exchange between photons and acoustic phonons, resulting in an optical frequency shift of the scattered light. This shift is proportional to the speed of sound in the material, and consequently to the longitudinal elastic modulus (M). However, it is unclear how Brillouin measurements, which characterize the mechanical response at GHz frequencies, relate to mechanical properties measured at much lower frequencies (~1 Hz) relevant to physiological conditions. Furthermore, as most biomaterials are hydrated, it remains unclear how the relative incompressibility of water influences the acoustic wave speed so as to affect Brillouin measurements of hydrated biomaterials.
In this study, we aim to establish the relationship between Brillouin frequency shift, acoustic wave speed and quasi-static elastic modulus of hydrogels of varying stiffness. Hydrogels are homogeneous and isotropic materials that mimic the poroelastic nature of biological tissues. Each measurement probes the mechanics of hydrogels in a significantly different frequency range: GHz for Brillouin imaging, MHz for ultrasound and Hz for unconfined compression tests. The acoustic wave speed falls into range from 1490 to 1533 m/s in both MHz (ultrasound) and GHz (Brillouin) frequency ranges. The quasi-static modulus correlates positively with Brillouin frequency shift, increasing from 6 to 54 kPa. All the results indicate the measurements obtained by Brillouin microscopy are capable of representing the material properties of hydrogels in quasi-static condition.
VIPA (Virtually Imaged Phased Array) based spectrometers are now routinely favoured over other types of spectrometers (such as scanning Fabry-Perot) for Brillouin imaging because VIPAs permit higher data acquisition speeds as compared to others. However, higher speeds mean lower photon counts at the camera used to acquire the spectra. The quality of optical components used is also important and have profound effect on the quality of the spectrum. Yet, these issues have not been addressed by various groups doing Brillouin imaging around the world. In this talk we examine the effect of the various optical components on the overall performance of the spectrometer both in one and two stage configuration. We define information content in the measured spectra and using information theoretic approach determine system parameters under various design conditions. We show for example, the spherical aberration imparted by the plano-convex cylinder lens usually placed at the entrance of the spectrometer reduces signal quality but it otherwise does not affect the accuracy of measurements. On the other hand, aberrations introduced by lenses further down the optical train may result in significant loss in localisation accuracy of the spectra. Our approach will aid users of VIPA based spectrometers designing better quality systems.