Dual-slope FD-NIRS during visual stimulation

Functional near-infrared spectroscopy (fNIRS) is a technique with a breadth of applications due to its ability to measure cerebral hemodynamics noninvasively and in naturalistic settings. Despite this, the field of fNIRS still struggles with contamination of the measured optical signals from superficial tissue hemodynamics. One possible approach is to collect optical data with sufficient spatial sampling information to allow for depth-resolved tomographic reconstructions. Alternatively, one may collect optical data that are intrinsically preferentially sensitive to the deeper, cerebral regions of the tissue. The dual-slope (DS) technique was created for this purpose, as it shows a great advantage over traditional single-distance (SD) techniques because of its greater relative sensitivity to brain vs. superficial extracerebral tissue. Along the same lines, phase data in FD-NIRS also features a greater depth sensitivity than intensity data in CW-NIRS.

Figure 1. (a) Visual stimulation protocol. (b) Folding average of ∆[HbO₂] and ∆[Hb] measured with FD-NIRS data (intensity, phase) in single-distance (SD) and dual-slope (DS) configurations, lowpass filtered to 0.2 Hz for visualization. (c) Functional activation image of the primary visual cortex obtained with single-distance phase data.

These advantages are demonstrated in Figure 1, which reports results of a visual stimulation protocol in a healthy 29 year-old male subject with a dual-slope FD-NIRS imaging probe placed over his occipital cortex (Fig. 1(a)). This protocol consists of an 8 Hz reversing checkerboard pattern displayed for 15 s, followed by 30 s of rest. During visual stimulation, blood flow in the primary visual cortex increases as a result of neurovascular coupling and elicits an increase in the concentration of oxy-hemoglobin ([HbO₂]) and a decrease in the concentration of deoxy-hemoglobin ([Hb]). This is the fNIRS signature of brain activation. The visual stimulation protocol is repeated 11 times to obtain folding averages of optical measurements as shown in Fig. 1(b), which reports ∆[HbO₂] (red lines) and ∆[Hb] (blue lines) obtained with different FD-NIRS data sets. Intensity and phase data are measured at a single distance (SD) between source and detector (either 25 or 37 mm), and in a dual-slope (DS) configuration that uses data collected at both 25 and 37 mm. The greater sensitivity to cerebral tissue featured by phase vs. intensity data, and by DS vs. SD configurations, accounts for the greater amplitude of the activation hemodynamics measured with phase vs. intensity, and measured with DS vs. SD (see Fig. 1(b)). Finally, Fig. 1(c) reports the activation map (as represented by the difference between ∆[HbO₂] - ∆[Hb] during stimulation and at rest) obtained with single-distance phase data. This functional image is obtained with the source-detector array shown in Fig. 1(a).

There is a clear advantage offered by dual-slopes (vs. single-distance) and phase (vs. intensity) data for functional activation studies in the human brain. However, this advantage is balanced by a potentially greater noise of DS and phase measurements that needs to be properly taken into account in DS FD-NIRS measurements.