In biomedical optics there is a need for models of biological tissues for calibration of instruments, as well as system design and optimization. Frequently, static uniformly dispersed mircospheres and dyes embedded in a nonscattering medium are used to mimic scattering and absorption by particles found in tissue. However, the static dispersed sphere approximation poorly describes scattering in most biological tissues for several reasons. First, it ignores the effects of collagen fibers which are a major constituent of many tissues and are better approximated as rods than spheres. Additionally, tissues are inhomogeneous and contain impurities such as lipid droplets, which affect the overall properties of the turbid medium. Finally, important tissue dynamics such as blood flow and the cell cycle are omitted.
In order to account for the above tissue complexities during design and calibration of biomedical optics instruments, we pursue the concept of a living optical phantom. The idea is to create artificial tissue models using current tissue engineering techniques where collagen ìrodsî, lipid ìimpuritiesî, and tissue dynamics are included in a well-controlled manner. Using such an approach, it is possible to control the levels of different tissue constituents in a model and study the corresponding instrumentation response. We hypothesize that rigorous correlation of engineered tissue morphology with measured signals would more precisely define exactly to which tissue constituents and processes the instrument is sensitive, and that this information will ultimately improve predictive capability of the instrument.
This report demonstrates the type of information offered in living optical phantoms by extracting optical scattering properties from optical coherence tomography (OCT) images of cell-remodeled collagen matrices that mimic the arterial wall. Because the turbid nature of tissue determines what types of optical signals can be detected, measurements of tissue scattering properties are common to many biomedical optics applications. One such application is OCT, which is an imaging technology that quantifies light backscattered from the tissue cross-section. We recently developed an OCT based algorithm that extracts the optical scattering properties (scattering coefficient μs and effective anisotropy factor geff) of tissues by fitting OCT signals as a function of depth to a theoretical model [1]. We have used this technique to evaluate the optical properties of OCT images of the arterial wall, and found that the distribution of optical properties varies between normal vessels and 3 different gross morphologies of atherosclerotic plaques. We now use our method to study how remodeling of collagen matrix by vascular smooth muscle cells in vitro affects their optical properties.