Introduction: Near-infrared spectroscopy (NIRS) and diffuse optical imaging (DOI), in the wavelength range 600-1000 nm, have been used for non-invasive optical studies of biological tissues for a long time. Some notable applications include tissue oximetry, pulse oximetry, assessment of blood flow and oxygen consumption in skeletal muscles, functional brain imaging, and optical mammography. Besides continuous-wave (CW) methods that use constant illumination, time-resolved methods either in the time domain (TD: pulsed illumination and time-resolved detection) or in the frequency domain (FD: intensity-modulated illumination and phase-resolved detection) have been introduced. Furthermore, slope methods based on the collection of data at multiple source-detector distances have been proposed, especially with CW and FD methods, to perform absolute measurements of tissue optical properties or to minimize sensitivity to superficial tissue layers. Slope techniques are typically based on either a single source (and multiple detectors) or a single detector (and multiple sources), in which case they may be termed “single-slope” methods. A “dual-slope” approach, identified as “self-calibrating,” was introduced to perform slope measurements that are insensitive to instrumental and optical coupling effects, resulting in calibration-free measurements.
The motivation of this work is to achieve quantitative optical measurements and preferential sensitivity to deep tissue using frequency-domain NIRS (FD-NIRS) in dual-slope configurations. This is important for non-invasive optical measurements to achieve preferential sensitivity to brain and muscle tissue underneath scalp/skull and skin/adipose layers, respectively.
Methods: Theoretical calculations based on diffusion theory were first run to characterize the spatial region of sensitivity of the dual slope technique implemented with FD-NIRS and two illumination points and two collection points that realize source-detector distances of 2.5 and 3.5 cm. Homogeneous and heterogeneous media were considered, with special emphasis on two-layered media. In vivo results were then obtained on human subjects from the primary visual cortex during visual stimulation, and from the forearm muscle during either venous occlusion or arterial occlusion in the upper arm.
Results: In both theoretical simulations and in vivo measurements with FD-NIRS, we consistently found enhanced depth sensitivity using phase vs. intensity data, and using dual-slope vs. single-distance data. We also found that the relative scattering properties of superficial and deeper tissue affect the depth sensitivity achieved by different optical measurements. In the case of brain measurements, we observed the lowest sensitivity to cortical hemodynamics using single-distance intensity, intermediate sensitivity using single-distance phase or dual-slope intensity, and maximal sensitivity using dual-slope phase. In the case of muscle measurements, the different hemodynamics and oxygen metabolic rates in superficial adipose tissue and deeper muscle tissue result in quantitatively and qualitatively different dynamics observed with different data types.
Conclusions: Dual-slope measurements feature desirable aspects of practical and conceptual significance that can help advance a number of spectroscopy and imaging applications in the field of non-invasive diffuse optics. Specifically, they can provide more specific measurements of cerebral hemodynamics in functional brain imaging, and more detailed characterization of skeletal muscle hemodynamics and oxygenation during vascular occlusion and exercise protocols.