Characterization of penetration depth as a function of optical fiber separation at various absorption and scatter coefficients for a noninvasive metabolic sensor

Abstract

A visible-near IR (500-1,000nm) fiber optic sensor is under development that is intended to non-invasively assess muscle metabolism through the measurement of tissue pH and oxygen partial pressure. These parameters are calculated from the spectra of hemoglobin and myoglobin in muscle. The sensor consists of transmit (illumination) fibers and receive (detection) fibers that are coupled to a spectrometer. Light from the probe must penetrate below the surface of the skin and into a 5-10mm thick layer of muscle. A study was conducted to quantify the relationship between transmit and receive fiber separation and sensor penetration depth below the surface of the skin. A liquid phantom was created to replicate the absorption (μa) and reduced scatter coefficient (μs') profiles typically found in human blood and tissue. The phantom consisted of a solution of Intralipid and India ink in the appropriate concentrations to achieve desired reduced scatter coefficient and absorption profiles. The reduced scatter coefficient of the liquid phantom was achieved to an accuracy of +/-10% compared to previously published data. A fixed illumination fiber and translatable detector fiber were placed in the liquid phantom, and the fiber separation was varied from 3-40mm. Values of μa and μs' varied from 0.03-0.40 cm-1 and 5.0-15.0 cm-1 respectively. Results from the experiment demonstrate a strong correlation between penetration depth and fiber separation. Additionally, it was found that penetration depth was not substantially influenced by absorption and scatter concentration. As signal-to-noise is an important parameter in many non-invasive biomedical applications, the relative signal as a function of fiber separation was determined to follow an exponential relationship.