Tissue near-infrared (NIR) spectroscopy is now more than 30 years old. The early promise of the technique as a unique tool for the noninvasive measurement of brain and muscle oxygenation is even now only beginning to be fulfilled. Despite the efforts of many hundreds of researchers, the basic problems associated with quantification and localization of the detected signal remain severe hindrances to the widespread clinical adoption of this technique. In contrast to its ubiquitous relative the pulse oximeter, the clinical cerebral and muscle oximeter is still rarely used outside the specialist research laboratory. A number of factors give rise to this situation:
• To be of immediate clinical utility, a cerebral oximeter should provide an absolute measurement of the hemoglobin oxygenation in the tissue of interest. Such information is readily available if blood samples are drawn from a patient and analyzed in a co-oximeter. Similarly, the pulse oximeter can provide this information for hemoglobin located in the arterial vascular bed. Unfortunately, providing a measurement of the average hemoglobin oxygenation in a thick tissue section such as the brain is fraught with complexities—so many, in fact, that early commercial cerebral oximeters did not attempt to measure this quantity at all. Instead, these devices measured only the change in tissue oxygenation, with reference to an unknown starting point. Such measurements can be of use to the physiologist, but are of relatively little use to, say, the neurologist who is running a head injury unit. Although some absolute hemodynamic variables such as absolute cerebral blood volume and flow can be estimated using such “trend” measurements, the procedure is time-consuming, is cumbersome, and requires both skilled operators and relatively sophisticated data interpretation.
• The head and other organs are physically heterogeneous, and it is not obvious which of several physiologically distinct regions contributes the majority of the measured signal. The most obvious example is the adult head, where the tissue of interest, the brain, must be interrogated through overlying tissues (the scalp, skull, and meninges) that are of considerable thickness and whose blood supplies are largely independent from the brain’s own blood supply. While modern instrumentation can successfully transilluminate the neonatal head, this is generally not possible with the much larger adult head. Thus, measurements must be made in a “reflection” mode, with source–detector spacing not exceeding approximately 5.0 cm. A naïve analysis then suggests that the interrogating light has sampled these overlying tissues to a similar or greater extent than the cerebral tissues. In consequence, there has been a long standing controversy about whether the oxygenation changes detected by NIR oximeters are truly reflective of changes occurring within the cerebral gray and white matter, particularly when used on adults.
It is the goal of this chapter to review the subject of quantification and localization of oxygenation measurements in tissue near-infrared spectroscopy. Considerable progress has been made in the last 10 years in instrumentation, data analysis, and theoretical modeling of light propagation in biological tissues. A number of instruments are now available that are capable of measuring absolute hemoglobin oxygenation, several of them commercially. New theoretical models have been developed that, given accurate knowledge of the optical properties of various tissues, can predict which regions of a heterogeneous organ contribute to the measured signal. New application areas have emerged - in particular, the exciting new field of functional near-infrared studies of the brain. Finally, the ultimate goal of localized spectroscopy - namely, the generation of oxygenation images - now appears to be a less remote possibility than before, with encouraging results having recently been obtained on tissue phantoms.
Online access to SPIE eBooks is limited to subscribing institutions.