Brain specificity of diffuse optical imaging: improvements from superficial signal regression and tomography

Abstract

Functional near infrared spectroscopy (fNIRS) is a portable monitor of cerebral hemodynamics with wide clinical potential. However, in fNIRS, the vascular signal from the brain is often obscured by vascular signals present in the scalp and skull. In this paper, we evaluate two methods for improving in vivo data from adult human subjects through the use of high-density diffuse optical tomography (DOT). First, we test whether we can extend superficial regression methods (which utilize the multiple source–detector pair separations) from sparse optode arrays to application with DOT imaging arrays. In order to accomplish this goal, we modify the method to remove physiological artifacts from deeper sampling channels using an average of shallow measurements. Second, DOT provides three-dimensional image reconstructions and should explicitly separate different tissue layers. We test whether DOT's depth-sectioning can completely remove superficial physiological artifacts. Herein, we assess improvements in signal quality and reproducibility due to these methods using a well-characterized visual paradigm and our high-density DOT system. Both approaches remove noise from the data, resulting in cleaner imaging and more consistent hemodynamic responses. Additionally, the two methods act synergistically, with greater improvements when the approaches are used together.

A review on continuous wave functional near-infrared spectroscopy and imaging instrumentation and methodology

“…This method was successfully tested using fNIRI data from adults (Biallas et al, 2012a;Saager and Berger, 2008;Saager et al, 2011) and newborns (Biallas et al, 2012b;Liao et al, 2010). A similar regression method ('superficial signal regression'), which uses the averages of all short (1.3 cm) SD channels as a regressor, was developed by Gregg et al (2010). In Saager et al (2011) an improved version of the approach was presented using an 8-channel probe configuration and 2 small SD channels (5 mm).…”

Separation of superficial and cerebral hemodynamics using a single distance time-domain NIRS measurement

“…However, due to the fundamental principles of light propagation in tissue, signals measured by fNIRS exhibit a higher sensitivity to absorption changes in superficial tissues compared to absorption changes in the deeper lying brain compartment [3,4]. Therefore light absorption changes resulting from spontaneous as well as task-evoked changes in the scalp perfusion [5][6][7] can mask cerebral signals and may cause false-positive results in functional studies [8]. Thus, a reliable separation of superficial and cerebral hemodynamic signals is required for better accuracy of fNIRS.…”

“…Examples are signal correction based on recording at two [9,10] or multiple [7] source-detector separations, application of principle or independent component analysis [11,12], adaptive filtering [13,14], general linear modeling (GLM) [6], wavelet coherence analysis [15] as well as methods exploring temporal correlation between oxy-and deoxyhemoglobin concentration changes induced by neuronal activation [16,17]. The latter approaches rely on assumptions about dynamic properties of extra-and intracerebral time series data.…”

Extended hierarchical Bayesian diffuse optical tomography for removing scalp artifact

“…It is empirically known that the hemodynamic change in the superficial layers, including the scalp and the skull, is not localized but instead spatially global [9][10][11][12][13]. To remove or isolate this global artifact, most methods have used temporal information, and often the temporal values of the short-distance measurement channel as a reference of superficial hemodynamic fluctuations for regression [9][10][11][12][13][14][15]. However, the inhomogeneity of the temporal patterns in the hemodynamic change of the superficial layers [16,17] or the temporal correlation between the hemodynamic responses in the scalp and in the cortex [7,8,18] negatively impacts such artifact-removal methods based on temporal information.…”